Available online at www.sciencedirect.

com

Carbohydrate Research 343 (2008) 1202–1211

Immobilization of glucoamylase by adsorption on carbon supports

and its application for heterogeneous hydrolysis of dextrin
Galina A. Kovalenko* and Larisa V. Perminova
Boreskov Institute of Catalysis, 630090 Novosibirsk, Russia
Received 7 December 2007; received in revised form 1 February 2008; accepted 5 February 2008
Available online 13 February 2008

Abstract—Glucoamylase (GA) was immobilized by adsorption on carbon support: on Sibunit, on bulk catalytic filamentous carbon
(bulk CFC) and on activated carbon (AC). This was used to prepare heterogeneous biocatalysts for the hydrolysis of starch dextrin.
The effect of the texture characteristics and chemical properties of the support surface on the enhancement of the thermal stability
of the immobilized enzyme was studied, and the rates of the biocatalyst’s thermal inactivation at 65–80 °C were determined. The
thermal stability of glucoamylase immobilized on different carbon supports was found to increase by 2–3 orders of magnitude in
comparison with the soluble enzyme, and decrease in the following order: GA on Sibunit > GA on bulk CFC > GA on AC. The
presence of the substrate (dextrin) was found to have a significant stabilizing effect. The thermal stability of the immobilized enzyme
was found to increase linearly when the concentration of dextrin was increased from 10 wt/vol % to 50 wt/vol %. The total stabil-
ization effect for glucoamylase immobilized on Sibunit in concentrated dextrin solutions was about 105 in comparison with the
enzyme in a buffer solution. The developed biocatalyst, ‘Glucoamylase on Sibunit’ was found to have high operational stability
during the continuous hydrolysis of 30–35 wt/vol % dextrin at 60 °C, its inactivation half-time (t1/2) exceeding 350 h. To improve
the starch saccharification productivity, an immersed vortex reactor (IVR) was designed and tested in the heterogeneous process
with the biocatalyst ‘Glucoamylase on Sibunit’. The dextrin hydrolysis rate, as well as the process productivity in the vortex reactor,
was found to increase by a factor of 1.2–1.5 in comparison with the packed-bed reactor.
Ó 2008 Elsevier Ltd. All rights reserved.

Keywords: Glucoamylase; Carbon supports; Adsorption; Heterogeneous biocatalyst; Thermal stability; Dextrin hydrolysis

1. Introduction tivation half-time (t1/2) 30–120 days, which corresponds

to a 3–12 month operation of the reactor without
A heterogeneous biocatalyst prepared by the immobili- reloading.1 A high value of t1/2 (P50–100 days) is essen-
zation of glucoamylase (1,4-a-D -glucan-glucohydrolase, tial to reduce the biocatalyst cost and to increase biocat-
EC 3.2.1.3) and designed for one of the key stages of alyst productivity up to a recommended value from 100
starch conversion—saccharification (hydrolysis of dex- to 10,000 kg of substrate converted per kg of biocata-
trin)—can become the basis for the development of a lyst.2 The most important characteristic for the indus-
new modern technology for production of sweeteners trial application of a biocatalyst is its high thermal
(treacle with variable hydrocarbon composition and stability at a pasteurization temperature of 60 °C or
glucose syrups) from renewable vegetable feedstock. higher. In the 1970s Corning Glass Co. carried out pilot
For successful commercialization of the heterogeneous plant tests of a packed-bed reactor with a glucose pro-
process of starch saccharification, a biocatalyst must ductivity of 450 kg per day. The reactor was filled with
convert 45% of the substrate in 15–20 min with an inac- a biocatalyst prepared by covalent binding of glucoam-
ylase onto the surface of macroporous silica.3 Inactiva-
tion half-times of this biocatalyst at 60 °C and 70 °C
* Corresponding author. Tel.: +7 383 3269743; fax: +7 383 3308056; were 150 h and 7.5 h, respectively.3 According to the
e-mail: galina@catalysis.ru opinion of the specialists working on this project, it is

0008-6215/$ - see front matter Ó 2008 Elsevier Ltd. All rights reserved.
doi:10.1016/j.carres.2008.02.006
 G. A. Kovalenko, L. V. Perminova / Carbohydrate Research 343 (2008) 1202–1211 1203

the low thermal stability of the immobilized gluco- studied. The operational stability of the developed bio-
amylase that was the main reason why this process catalyst ‘Glucoamylase on Sibunit’ under model techno-
was not commercialized. It is important to mention that logical conditions of starch saccharification was
no biocatalyst based on immobilized glucoamylase with determined. An immersed vortex reactor (IVR) of a
satisfactory thermal stability at P60 °C has been new type was designed and tested in starch saccharifica-
developed to date. The best result obtained so far is tion by the biocatalyst ‘Glucoamylase on Sibunit’. A
the inactivation half-time 300 h at 50 °C obtained for lab-scale scheme for starch bioconversion was developed
a biocatalyst prepared by immobilization of GA on and tested.
polystyrene.4
Another factor limiting the commercialization of
heterogeneous biocatalysis is the relatively low produc- 2. Experimental
tivity of packed-bed reactors, particularly in processes
controlled by the substrate diffusion to the immobilized 2.1. Materials
enzyme. The results of the studies on the development of
new types of bioreactors may be used as the basis for the 2.1.1. Enzyme and substrate. Glucoamylase from a
modernization of the existing technology or develop- commercial sample of GlucoawamorinÒ (‘Sibbio-
ment of new competitive ones. The most important pharm’, Novosibirsk, Russia) was used for adsorptive
problem that has to be solved by the new reactor design immobilization on carbon supports. The specific gluco-
is the significant decrease of diffusion limitation by pro- amylase activity at 50 °C and pH 4.6 was equal to
viding intensive forced mass transfer of the substrate to 200 U/mg of protein.
the biocatalyst. This mass transfer can be intensified, for The dextrin used as a substrate for soluble and immo-
example, by the creation of a vortex movement in the bilized glucoamylase was prepared as follows: Dry pota-
liquid. The idea to use Taylor–Poiseulle vortex flow to, corn, or wheat starch was mixed with a dry enzyme
for carrying out heterogeneous biocatalytic processes sample of a-amylase, AmylosubtilinÒ (‘Sibbiopharm’,
with immobilized enzymes and the theoretical back- Novosibirsk, Russia), and carefully homogenized. The
ground for the construction of such bioreactors was sug- mixture thus obtained was suspended in distilled water,
gested by Iosilevskii.5 Later this idea was implemented heated under intensive stirring to 85 °C, and maintained
in a vortex flow reactor that consisted of two cylinders, at this temperature for 15–20 min. At the end of this
the inner cylinder being immersed into the substrate time the mixture was boiled for 5 min to inactivate the
solution and rotated in it.6–8 a-amylase, then cooled to room temperature, and
Earlier we have studied the biocatalytic properties of filtered. The concentration of dry-weight substances in
glucoamylase immobilized on macrostructured and the substrate solutions (wt/vol %) was measured using
granular carbon-containing alumina–silica and alumina an RL3 refractometer (‘PZO Warszawa’, Poland).
supports differing by the morphology of the carbon The molecular weight of the prepared dextrin was
layer synthesized on the surface.9,10 We have shown that estimated at 3–5 kDa. The dextrose equivalent (DE)
the thermal stability of glucoamylase adsorbed on a of dextrin substrate was determined to be equal to ca.
honeycomb monolith coated by carbon nanofibers 10.
increases by a factor of 20 in comparison with the solu-
ble enzyme; the inactivation half-time of this biocatalyst 2.1.2. Supports. SibunitÒ and bulk CFC were devel-
at 65 °C was 9.5 h.9 Earlier we had studied the biocata- oped at the Boreskov Institute of Catalysis as supports
lytic properties of glucoamylase immobilized on granu- and adsorbents with desired regulated properties.12–14
lar pure carbon supports, including Sibunit, graphite, These supports were earlier studied for the immobiliza-
and sapropel.10,11 Mesoporous Sibunit was found to tion of some biologically active substances (amino acids,
be the most efficient adsorbent for glucoamylase.10 In proteins, and non-growing bacterial cells).15,16 Activated
the present study, we have continued a comparative carbon (AC) is a commercial product that is used as an
investigation of granular pure carbon supports for the adsorbent in the food industry for the decolorization of
adsorptive immobilization of enzymes. sugar syrups and for drinking water purification. The
The present study was devoted to the investigation of physical characteristics of the supports in the study are
the properties (activity and stability) of biocatalysts shown in Table 1.
prepared by glucoamylase adsorptive immobilization
on granular pure carbon supports: Sibunit, bulk cata- 2.2. Methods
lytic filamentous carbon (bulk CFC), as well as activated
carbon (AC). The constants of thermal inactivation of 2.2.1. Method of enzyme immobilization. The immobili-
soluble and immobilized glucoamylase were determined zation of glucoamylase on carbon supports was carried
and compared. The effect of the substrate (dextrin) on out by adsorption under static conditions at 18–22 °C
the thermal stability of immobilized glucoamylase was as described earlier.10,11
1204 G. A. Kovalenko, L. V. Perminova / Carbohydrate Research 343 (2008) 1202–1211

Table 1. Texture characteristics of the carbon supports

Support SSP BET, m2/ga SSP, m2/g VR pore Predominant pore Accessible surface Accessible pore
(without micropores)b (without micropores) mL/gb diameter, nmb area SACC, m2/gc volume, mL/gc
Sibunit 555 95 0.86 18 92 0.85
Bulk CFC 162 67 0.38 11 42 0.34
AC 1296 49 0.53 4 9 0.47
a
Determined by BET method using thermal desorption of argon.
b
Determined by mercury porosimetry.
c
Calculated from pore size distributions data for pores larger than 10 nm.

2.2.2. Characterization of supports. The specific surface the following types: (1) A differential gradientless reac-
areas of the carbon supports were measured by the ther- tor was used to determine the thermal inactivation
mal desorption of argon using a Sorbi-M instrument kinetic constants. Due to a thin bed, the degree of
(‘Meta’, Russia). The pore size distributions were substrate conversion during one pass through the bio-
determined by mercury porosimetry using an Auto-Pore catalyst did not exceed 0.1%. As a result there were no
9200 instrument (‘Micromeritics’, USA). The texture gradients of the product and substrate concentrations
characteristics of the carbon supports are reported in along the biocatalyst bed. (2) A packed-bed reactor
Table 1. was used to determine the stability of the biocatalysts.
The surface acid–base properties of the adsorbents The height of the biocatalyst bed was varied from 0.5
used were studied by potentiometric titration at 20– to 5.7 cm. Reactors (1) and (2) were a temperature-
22 °C using acetic (pKa 4.75) and hydrochloric acids, controlled glass column with a length of 10-cm and a
and ammonium (pKb 4.75) and sodium hydroxides as diameter of 1-cm filled with a biocatalyst. The substrate
indicators for measuring basicity and acidity, respec- solution was circulated through the biocatalyst bed
tively. Support samples with surface areas of 50– using a peristaltic pump with a flow rate that was varied
100 m2 were immersed into the indicator solutions and from 3 to 550 mL/min. (3) An immersed vortex reactor
maintained under stirring for 30 min to obtain station- (IVR) was used to test this new reactor type in dextrin
ary pH values. The calculations of concentrations of hydrolysis. Its construction and principle of operation
the acid/base surface sites expressed in nmol/m2 were are described in an article from this laboratory.17 The
performed in the monolayer coverage region corres- rotation rate (x) of the reactor body filled with a biocat-
ponding to the plateau on the Langmuir isotherms alyst was varied from 50 rpm to 1500 rpm. A biocatalyst
obtained. with 2–4 mm granules was placed in the lower plate of
The hydrophobic properties were evaluated by naph- the body, and the upper plate was screwed on its top.
thalene adsorption from its saturated solution at 20– For finely dispersed supports a toroidal Capron filter
22 °C. Support samples with 1 m2 surface area were was prepared, filled with the biocatalysts, and placed
degassed using a water-jet pump to improve their wetta- inside the reactor body. Then, the body was immersed
bility, then put in contact with 0.09 mM naphthalene into a flask filled with 300–500 mL of the substrate
solution for 60 min. The naphthalene concentrations solution. Its temperature was controlled at 50 ± 5 °C.
were measured spectrophotometrically at 220 nm and Specially designed baffles were used to prevent the
were calculated using extinction coefficient 1.32  formation of a cone crater in the twisted liquid. The sub-
105 L cm1 mol1. The amounts of adsorbed naphtha- strate solution was sucked trough the bottom outlet in
lene were determined from the decrease of its concentra- the reactor body and was circulated through the biocat-
tion and expressed in nmol/m2. alyst bed multiple times for 1 h. The flow rate of the sub-
strate solution through the biocatalyst bed inside reactor
2.2.3. Measurements of glucoamylase activity. Stan- body increased proportional to x1/2. The reactor design
dard conditions commonly used for the measurement has been protected by a Russian patent.18
of the activity of soluble and immobilized glucoamylase
were the following: 50 °C, 0.05 M acetate buffer pH 4.6, 2.2.4. Measurements of stability of immobilized
and 10 w/vol % solution of potato dextrin as substrate. enzyme. The thermal stability of soluble glucoamylase
The amount of glucose (in lmol) generated for 1 min and its inactivation constant ðk Ein Þ were determined by
was used as the Unit of enzyme activity (U). The specific GlucoawamorinÒ (2.0 mg of protein/mL) heating at
enzyme activities were calculated as the amount of U per 60–80 °C in a buffer solution of pH 4.6. Aliquots were
mg of the protein (U/mg) and as the amount of U per g taken every 15 min and immediately cooled down in
of the dry-weight biocatalyst (U/g) for soluble and an ice bath. Then, the residual glucoamylase activity
immobilized GA, respectively. The dextrin hydrolysis was measured under the standard conditions at 50 °C
by immobilized GA was carried out using reactors of as described above.
 G. A. Kovalenko, L. V. Perminova / Carbohydrate Research 343 (2008) 1202–1211 1205

The thermal stability of immobilized glucoamylase pore structure of Sibunit and bulk CFC (Fig. 2a and
and its inactivation constants in a buffer without sub- b). Meanwhile, activated carbon has a bimodal pore
strate ðk Ein Þ and in a buffer with substrate ðk ES
in Þ were structure with the prevalence of macro- and micropores
determined as follows: A biocatalyst sample (0.1 g) (Fig. 2c). The accessible surface areas (SACC) of the car-
was heated at 65–80 °C directly in reactor (1) described bon supports were estimated using the pore size distribu-
above for 15–240 min in a circulated buffer solution of tion data and assuming that all pores larger than 10 nm
pH 4.6 without a substrate and in the presence of the are accessible for enzyme adsorption. The calculations
dextrin in the concentration varied from 1 to 53 wt/vol gave us the following order of the supports arranged
%. Then, the biocatalyst was washed with the cool buffer in decreasing accessible surface area: Sibunit (92 m2/
solution for 3–5 min, and its residual activity was just g) > CFC (42 m2/g) > AC (9 m2/g) (Table 1).
measured under the standard conditions at 50 °C as de- The carbon supports that were studied possessed pro-
scribed above. nounced basic properties, and oxygen-containing C@O
The operational stability of the biocatalyst prepared groups (including lactone) and phenolic hydroxyl groups
by the adsorptive immobilization of GlucoawamorinÒ were present on their surface.19,20 The total concentra-
on Sibunit (named ‘Glucoamylase on Sibunit’) was tion of basic sites on the surface was (1–2)  107 mol/
determined in a packed-bed reactor under conditions m2, and the fraction of strong basic sites did not exceed
simulating the technological regime of the starch sac- 30% (Table 2). No acid groups were observed on the sur-
charification process. A 32 wt/vol % solution of dextrin face of Sibunit and bulk CFC, whereas their concentra-
was flushed through the biocatalyst bed at 60 °C for the tion on AC was 20% of the total concentration of the
duration of an 8-h reaction cycle. The ratio of the bio- acidic and basic sites (Table 2). The hydrophobic proper-
catalyst weight to the substrate solution circulated was ties of the supports were similar (Table 2). The basicity of
1:150. For each cycle the residual glucoamylase activity the studied supports decreased in the following order:
was measured under the same conditions for the first bulk CFC (209 nmol/m2) > Sibunit (143 nmol/m2) >
hour of reactor operation. AC (72 nmol/m2) (Table 2).

3.2. Properties of glucoamylase immobilized on the

3. Results and discussion carbon supports

3.1. Characteristics of carbon supports Comparison of the properties of glucoamylase immobi-

lized on different carbon supports showed that the
The carbon supports in the study varied in surface mor- enzyme adsorption on the support enhanced propor-
phology (Fig. 1). As is seen from Figure 1a, the surface tionally to the accessible surface area (Table 3), so that
of Sibunit was formed by round coke deposits of pyro- a 2-fold increase of the accessible surface area of the
lytic carbon. As is seen from Figure 1b, the surface of support leads to a 2–3-fold increase of the enzyme
bulk CFC was formed by the chaotic interlacing of car- adsorption. Then, knowing the accessible surface area
bon nanofibers of 50–100 nm diameter and P1 lm of the supports (Table 1), as well as the adsorption val-
length. The resulting surface was very rough. As is seen ues and the molecular weight of glucoamylase
from Figure 1c, the surface of AC was formed by graph- (100 kDa), it was possible to calculate the apparent size
ite-like carbon, which gave a smoother surface. of an adsorbed enzyme molecule. Table 3 shows that
The analysis of the pore size distributions and specific when the amount of adsorbed enzyme is increased, the
surface areas of the carbon supports showed that mes- apparent size of the adsorbed enzyme molecules is de-
opores with a size of 20–100 nm, which are suitable creased, that is, the adsorption layer becomes denser.
for the immobilization of enzyme molecules with a crys- The calculations also showed that the densest layer of
tallographic size of almost 10 nm, predominated in the enzyme molecules was formed on the surface of

Figure 1. SEM images of the surface of studied carbon supports: (a) Sibunit; (b) bulk CFC; (c) activated carbon.
1206 G. A. Kovalenko, L. V. Perminova / Carbohydrate Research 343 (2008) 1202–1211

Figure 2. Pore size distribution diagrams of the carbon supports: (a) Sibunit; (b) bulk CFC; (c) activated carbon.

Table 2. Acid–base and hydrophobic properties of the carbon supports

Support Aciditya, nmol/m2 Basicitya, nmol/m2 Hydrophobicity, nmol/m2
Indicator 1 Indicator 2 Indicator 3 Indicator 4
Sibunit 0 0 41 143 184
CFC 0 0 74 209 161
AC 0 21 12 72 152
a
Indicators: 1, ammonium hydroxide, pKb 4.75; 2, sodium hydroxide; 3, acetic acid, pKa 4.75; 4, hydrochloric acid.

Table 3. Adsorption of glucoamylase on carbon supports and activity of immobilized enzyme

Support (granule size 0.2–1 mm) Adsorption, mg/g Diameter of adsorbed enzyme Specific activity, Biocatalyst activity,
molecule, nm (calculation) U/mg of protein U/g of dry biocatalyst
Sibunit 10.0 39 50.0 500
12.9 34 32.2 415
24.1 25 21.9 530
Bulk CFC 4.3 41 90.7 393
8.2 29 32.7 268
AC 4.6 18 18.3 84

activated carbon. In this case the surface area occupied AC support was lower than on the other studied sup-
by one molecule of adsorbed enzyme was close to the ports (Table 3). The surface areas occupied by one mol-
size of the hydrated enzyme molecule (20 nm). Mean- ecule of enzyme adsorbed on Sibunit and bulk CFC
while, the specific activity of the enzyme immobilized on were almost the same (Table 3). However, the enzymatic
 G. A. Kovalenko, L. V. Perminova / Carbohydrate Research 343 (2008) 1202–1211 1207

activity ran down when the enzyme adsorption in- GlucoawamorinÒ on Sibunit that has been named ‘Glu-
creased, and as a result, the adsorption layer became coamylase on Sibunit’.
denser (Table 3). Table 3 also shows that the looser
structure of the layer formed by the adsorbed protein 3.3. Properties of glucoamylase immobilized on Sibunit
molecules provides the higher specific activity of the
immobilized enzyme. These results seem to indicate that Note that the values of observed activity of the prepared
the enzyme active sites are deformed and/or partially heterogeneous biocatalysts and specific activity of the
blocked in a dense adsorption layer. immobilized enzymes reported in Table 3 were all mea-
Thermal inactivation of glucoamylase adsorbed on sured in the kinetic region for the dextrin hydrolysis
the carbon supports has been compared with that of reaction. To secure it, we selected the conditions for
the soluble enzyme. For the latter, the thermal inactiva- overcoming diffusion control of the reactions in different
tion constant ðk Ein Þ and inactivation half-time (t1/2) at types of reactors. For the packed-bed reactor the flow
65 °C were found to equal 3.3  102 min1 and rate of substrate solution through the biocatalyst bed
21 min, respectively. At temperatures P70 °C soluble was selected to make sure that the reaction rate did
glucoamylase was not stable and quickly became inacti- not depend on the flow rate; therefore, there were no
vated. Meanwhile, at 65 °C the value of k Ein for the external diffusion limitations. As is seen from Figure 3,
immobilized glucoamylase (Table 4) was 2–10 times the process of mass transfer of substrate (dextrin) to
higher than for the soluble enzyme. immobilized enzyme (glucoamylase) did not influence
Comparison of inactivation constants of immobilized on the hydrolysis rate at a flow rate more than 30 mL/
glucoamylase showed that the enzyme adsorbed on mes- min. It was found that the greater the height of the
oporous Sibunit has the highest stability (Table 4). Glu- bed in the reactor, the greater was the hydrodynamic
coamylase adsorbed on Sibunit preserved high resistance, and the greater was the flow rate to overcome
enzymatic activity even at 80 °C. At this temperature diffusion restrictions (Fig. 3). All subsequent studies
when it was immobilized on other carbon supports, were carried out at 20–30 mL/min.
the enzyme was inactivated relatively fast (Table 4).
The constants for thermal inactivation at 75 °C for
immobilized glucoamylase increased in the following
order: GA on Sibunit (1.9  102 min1) > GA on
bulk CFC (6.5  102 min1) > GA on AC (1.4 
101 min1). This order was analogous to and matches
the ones observed for the decrease of accessible surface
area of the support and the amount of mesopores in
the pore structure (Table 2). Thus, the above data
proved that the pore structure of the carbon supports
studied, rather than their hydrophobic–hydrophilic
properties, plays the key role in improving the stability
of the immobilized glucoamylase. Mesopores matching
the size of hydrated enzyme molecules and providing
multipoint enzyme binding with the surface of the sup-
port are required for the preparation of highly stable
biocatalysts.
Thus, these results suggest that mesoporous Subunit is Figure 3. Dependence of the observed biocatalyst activity in a packed-
the most efficient support for the preparation of a highly bed reactor on the circulatory flow rate of substrate solution through
the biocatalyst bed. (1) The biocatalyst is 0.5 cm (d); (2) the bed height
stable heterogeneous biocatalyst for starch saccharifica- are 2.0 cm (j) and 5.7 cm (M). The activity was measured under the
tion. All subsequent studies were carried out using bio- standard conditions. The biocatalyst was prepared by the adsorption
catalysts prepared by the adsorptive immobilization of of GA on small Sibunit granules (£ 0.2–1 mm).

Table 4. Thermal inactivation constants of glucoamylase immobilized on the carbon supports

T, °C Sibunit a = 10.0 mg/g Bulk CFC a = 8.2 mg/g Activated carbon a = 4.6 mg/g
k Ein , min1 t1/2 k Ein , min1 t1/2 k Ein , min1 t1/2
3 3 2
65 3.4  10 3.4 h 7.7  10 1.5 h 1.4  10 49 min
70 7.9  103 1.5 h 3.1  102 22 min 6.6  102 10 min
75 1.9  102 36 min 6.5  102 11 min 1.4  101 5 min
80 3.6  102 19 min Unstable Unstable
1208 G. A. Kovalenko, L. V. Perminova / Carbohydrate Research 343 (2008) 1202–1211

Then, the size of the support granules was selected in

such a way that the observed hydrolysis rate did not de-
pend on the granule size. This way the restriction by the
internal diffusion was overcome. For glucoamylase
immobilized on Sibunit, the dextrin hydrolysis rate
was found to decrease when the size of the support gran-
ules exceeded 1 mm (Fig. 4). Meanwhile, the activity of
the biocatalyst prepared using support granules smaller
than 1 mm was constant and relatively high (Fig. 4).
Hence, for the biocatalyst with the granule size larger
than 1 mm dextrin diffusion inside the pore space of
the granules was the rate-limiting stage of the starch sac-
charification process. The biocatalysts prepared using
small granules of the support worked in the kinetic re- Figure 5. Dependence of the initial specific activity of soluble and
gime without diffusion limitations. immobilized glucoamylase on the initial concentration of potato
Under the conditions for kinetic regime, the tempera- dextrin. (1) ‘Glucoamylase on Sibunit’. (2) GlucoawamorinÒ in
ture coefficients of dextrin hydrolysis were found to be solution. The activity was measured at 50 °C and pH 4.6.
the same equal to 1.6 for both immobilized glucoamy-
lase and the soluble enzyme. This was an additional con-
firmation of the fact that the biocatalyst ‘Glucoamylase 2.7  104 M (0.6 and 1.0 wt/vol %) for soluble and
on Sibunit’ worked in the kinetic region. immobilized glucoamylase, respectively.
The investigation of the reaction kinetics in the It was found that the optimum pH for the glucoamy-
Michaelis–Menten coordinates for soluble and immobi- lase activity did not change after immobilization on the
lized glucoamylase showed that the biocatalyst ‘Gluco- carbon support and remained at pH 4.5–5.0 (results not
amylase on Sibunit’ exhibited a high activity in a wide shown). The only difference was that the immobilized
range of dextrin concentrations from 10 to 35 wt/vol glucoamylase showed maximum activity in a narrower
%, whereas a more complicated kinetic dependence pH range, whereas for the enzyme in solution, the opti-
was obtained for the soluble enzyme (Fig. 5). As is seen mum pH range was pH 3.0–6.0 (results not shown). This
from Figure 5, after immobilization, glucoamylase re- fact can be explained by preferential adsorption of glu-
tained 20% of the activity of the soluble enzyme. The coamylase from the solutions of the GlucoawamorinÒ,
dissociation constant of the enzyme–substrate complex, which also contains ballast proteins and accompanying
KS, increased after immobilization. Assuming that the enzymes. Thus, glucoamylase homogeneity and its de-
average molecular weight of dextrin is 3–5 kDa, these gree of purification improve after adsorption. Indeed,
constants were estimated to be 1.7  104 M and the investigation showed that the specific enzyme activ-
ity in the solution over the Sibunit support (after
adsorption) was 1.5 times lower than the specific activ-
ity before adsorption, indicating preferential adsorption
of glucoamylase molecules. Glucoamylase in a relatively
pure enzyme sample from different companies (‘BDH’,
‘Sigma’, and ‘Serva’) is known to have a relatively nar-
row optimum range of pH 4.5–4.6.
The investigation of the temperature optimum for the
dextrin hydrolysis reaction showed that glucoamylase
both in solution and in the immobilized state has a rel-
atively narrow temperature optimum, showing the high-
est activity at 65–70 °C (results not shown).
As mentioned above, the most important characteris-
tic for the industrial application of a biocatalyst is its
stability. The glucoamylase thermal stability increased
by more than an order of magnitude after its immobili-
zation by adsorption on the surface of the carbon sup-
ports evaluated (Table 3). Investigation of the thermal
inactivation of immobilized glucoamylase at 80 °C
Figure 4. Dependence of the biocatalyst activity on the Sibunit
support granules size (mm). The reaction was carried out under the showed that in the presence of the substrate (dextrin)
standard conditions. The rate of the substrate solution circulation the biocatalyst ‘Glucoamylase on Sibunit’ was substan-
through the biocatalyst bed was 30 mL/min. tially (by an order of magnitude) more stable than in
 G. A. Kovalenko, L. V. Perminova / Carbohydrate Research 343 (2008) 1202–1211 1209

the buffer solution. For example, in the 37 wt/vol %

dextrin solution, the inactivation constant of the
immobilized glucoamylase at 80 °C was equal to
3  103 min1, whereas in the buffer solution it was
equal to 4.0  102 min1 (Fig. 6). The increase of the
substrate concentration from 1 to 50 wt/vol % resulted
in nearly linear growth of the thermal stability of the
‘Glucoamylase on Sibunit’ (Fig. 7). As is seen from Fig-
ure 7, in the 53 wt/vol % dextrin solution, the immobi-
lized glucoamylase completely preserved its initial
activity. Further increase of the concentration of dextrin
prepared by starch hydrolysis is practically impossible
because it is difficult to carry out the liquefaction of a
starch suspension with the concentration of dry-weight
substances more than 50 wt/vol %.
Figure 8. Dependence of the biocatalyst activity on operation time in
The investigation of the operational stability of the
the starch saccharification process (operational stability of the biocat-
‘Glucoamylase on Sibunit’ showed stepwise inactivation alyst). The activity was tested under the following conditions: 60 °C;
of the biocatalyst under conditions simulating the starch 32 wt/vol % solution of potato dextrin; the rate of the substrate flow
through the biocatalyst bed was 20 mL/min.

saccharification technological process (60 °C, 32 wt/vol

% dextrin) (Fig. 8). The biocatalyst inactivation half-
time under these conditions exceeded 350 h. A compar-
ison of the characteristics of the biocatalyst prepared in
this study with the results known from the literature
showed that the main parameters of the biocatalyst
‘Glucoamylase on Sibunit’ exceeded those of the best
reported analog4 as well as that of the Corning Glass
biocatalyst tested on the pilot plant scale3 (Table 5).
The calculations show that the total productivity
of the ‘Glucoamylase on Sibunit’ biocatalyst after
operation for 2  t1/2 (700 h) at a productivity of
7:6 kg h1 kg1
cat in a novel type of immersed vortex reac-
tor will be 5.3 tons of glucose per kg of the biocatalyst.
Figure 6. Changes in the activity of immobilized glucoamylase at
80 °C: (1) Buffer solution; (2) 37 wt/vol % solution of potato dextrin.
We have also compared different types of reactors,
The residual activity was measured under the standard conditions. including an immersed vortex reactor (IVR), in the pro-
cess of dextrin hydrolysis. As mentioned above, the
main attention in the IVR design was devoted to the
substantial intensification of mass transfer of the sub-
strate from the liquid phase to the enzyme immobilized
on a solid porous support and elimination of stagnant
zones. The laboratory model of the developed immersed
vortex reactor was tested in heterogeneous dextrin
hydrolysis by the biocatalyst ‘Glucoamylase on Sibunit’,
and optimal operational conditions were determined for
this reactor. It was shown that the observed dextrin
hydrolysis rates practically did not change when the
reactor body’s rotation rate was increased above
300 rpm (Fig. 9). It means that at this rotation rate the
external diffusion limitations were minimized. It was
also found that higher body rotation rates were required
to overcome the external diffusion limitations for the
Figure 7. Dependence of the residual GA activity after heating the higher active biocatalyst. For instance, the body rota-
biocatalyst at 80 °C for 2 h on the dextrin concentration. The residual tion rates were higher by 100–200 rpm for the more
activity was measured under the standard conditions. active biocatalysts than for the less active ones
1210 G. A. Kovalenko, L. V. Perminova / Carbohydrate Research 343 (2008) 1202–1211

Table 5. Comparative characteristics of the biocatalyst ‘Glucoamylase on Sibunit’ with literature data
Properties of the biocatalyst ‘Glucoamylase on Sibunit’ ‘Glucoamylase on polystyrene’4 CORNING GLASS biocatalyst
‘Glucoamylase on silica’3
Inactivation half-time, h 350 (60 °C) 300 (50 °C) 150 (60 °C)
7.5 (70 °C)
Activity, lmol min1 g1 700 500 12
Specific enzyme activity, lmol min1 mg1 60 15 110
Reactor Vortex Fluidized-bed Packed-bed

Figure 9. Effect of IVR body rotation rate on the observed activity of the biocatalyst ‘Glucoamylase ion Sibunit’: (1) Granules with diameter £ = 2–
3 mm; (2) granules with diameter £ = 0.2–1 mm. The activity was measured under the standard conditions.

(Fig. 9). As noted above, the dextrin hydrolysis process exchange; (4) enzymatic hydrolysis of dextrin by gluco-
was carried out deeply in the internal diffusion region. amylase at 55–60 °C and pH 4.5 for 24–96 h (homoge-
So, the reaction rate observed in IVR increased by an neous stage); (5) enzyme inactivation and extraction.
order of magnitude when the size of Sibunit granules The lab-scale results obtained in this study suggest the
was decreased from 3 to 0.2 mm (Fig. 9). following changes to this technology: Stages (1) and
The efficiency of the IVR operation in dextrin hydro- (2) can be combined due to the high thermal stability
lysis was estimated by comparing the reaction rates of the used a-amylase from AmylosubtilinÒ, and hydro-
observed in a vortex reactor and in a packed-bed reactor lysis can be carried out at higher temperature 90 °C in
(PBR) under conditions with minimum diffusion limita- aqueous solutions at pH 5.5–6.0. Stage (4) can be carried
tions. The most active biocatalyst ‘Glucoamylase on out in a heterogeneous regime using the biocatalyst
Sibunit’ prepared by glucoamylase adsorption on meso- ‘Glucoamylase on Sibunit’ and the immersed vortex
porous fine-disperse Sibunit was used in these experi- reactor in 2–6 h (instead of 24–96 h). It should be noted
ments. It was found that in PBR the density of the that the carbohydrate composition of starch treacle can
biocatalyst bed and the hydrodynamic resistance in- be easily regulated by simple stopping of the reactor and
creased strongly when the bed height (h > 6 cm) and its removal from the reaction mixture. Due to the multi-
the circulation rate (>200 mL/min) were increased. ple use of glucoamylase and consequent lack of protein
The comparison of the dextrin hydrolysis rates in kinetic impurities, some steps of the final product purification in
region for the two types of reactors showed that the stages (3) and (5) are significantly simplified.
reaction rate and, as a consequence, the reactor produc-
tivity were higher by 20–50% in the vortex reactor than 3.4. Conclusions
in the packed-bed reactor (compare maximal observed
biocatalytical activity in Figs. 9 and 3). The data prove that the pore structure of the carbon
The modern technology used for starch conversion supports that were studied, rather than their hydropho-
into food sweeteners (treacle, glucose syrups) is based bic–hydrophilic properties, plays the key role in improv-
on enzymatic (rather than acid) hydrolysis2 and includes ing the stability of the immobilized glucoamylase.
the following stages: (1) gelatinization of 40% starch Mesopores matching the size of hydrated enzyme mole-
suspension at 100–110 °C; (2) hydrolysis by the cules and providing multipoint enzyme binding with the
enzymes, a-amylase and pullunase at 60 °C and pH 6.5 support are required for the preparation of highly stable
for 10–100 min (homogeneous stage); (3) filtration, ion heterogeneous biocatalysts.
 G. A. Kovalenko, L. V. Perminova / Carbohydrate Research 343 (2008) 1202–1211 1211

It has been shown that after the immobilization of Sibunit and bulk CFC supports, respectively. The
glucoamylase from GlucoawamorinÒ on the mesopor- authors acknowledge Sergey V. Sukhinin for the inven-
ous carbon support, Sibunit, some kinetic characteristics tion of a vortex reactor. The researches were partially
of the enzyme, including the constant of enzyme affinity supported by the Grant 4.4 of the Siberian Branch of
to the substrate (dextrin with molecular weight 3– the Russian Academy of Science.
5 kDa), pH- and temperature optima of the enzymatic
reaction are practically unchanged. Meanwhile, the ther-
mal stability of the enzyme considerably improves in
References
comparison with the enzyme in solution. It has been
found that the thermal stability of the immobilized 1. Daniels, M. J. In Process Engineering Aspects of Immobi-
enzyme is significantly affected by the substrate, so that lised Cell Systems; Webb, C., Black, G. M., Atkinson, B.,
an increase of the dextrin concentration from 1 to 50 wt/ Eds.; Institution of Chemical Engineers: Rugby, War-
vol % results in a linear growth of the thermal stability wickshire, 1986; Chapter 13, pp 218–224.
of the immobilized glucoamylase. It has been found that 2. Bucholdz, K.; Kasche, V.; Bornscheuer, U. T. Biocata-
lysts and Enzyme Technology; Wiley-VCH: Weinheim,
the thermal stability of the immobilized enzyme in a 2005.
53 wt/vol % dextrin solution is P105 times better than 3. Reily, P. J. Appl. Biochem. Bioeng. 1979, 185–207.
that of the enzyme in solution. The operational stability 4. Bahar, T.; Celebi, S. S. Enzyme Microb. Technol. 2000, 26,
of the biocatalyst ‘Glucoamylase on Sibunit’ prepared 33–38.
by GlucoawamorinÒ immobilization on Sibunit is rela- 5. Iosilevskii, G.; Brenner, H.; Moore, C. M. V.; Cooney, C.
L. Philos. Trans. R. Soc. London, Ser. A 1993, 45, 259–
tively high. Under model technological conditions of 294.
dextrin hydrolysis at 32 wt/vol % concentration at 6. Giordano, R. C.; Giordano, R. L. C.; Prazeres, D. M. F.;
60 °C, the inactivation time exceeded 350 h (30 working Cooney, C. L. Chem. Eng. Sci. 1998, 53, 3635–3652.
days). 7. Giordano, R. L. C.; Giordano, R. C.; Prazeres, D. M. F.;
The immersed vortex reactor (IVR) suggested in this Cooney, C. L. Chem. Eng. Sci. 2000, 55, 3611–3626.
8. Giordano, R. L. C.; Giordano, R. C.; Cooney, C. L.
study significantly improves the productivity of hetero- Process Biochem. 2000, 36, 1093–1101.
geneous dextrin hydrolysis, which is controlled both 9. Kovalenko, G. A.; Komova, O. V.; Simakov, A. V.;
by the internal and the external diffusion of the substrate Khomov, V. V.; Rudina, N. A. J. Mol. Catal. A: Chem.
to the immobilized glucoamylase. The external diffusion 2002, 182–183, 73–80.
limitations were completely overcome at a reactor body 10. Kovalenko, G. A.; Perminova, L. V.; Plaksin, G. V.;
Chuenko, T. V.; Komova, O. V.; Rudina, N. A. Appl.
rotation rate above 300 rpm, whereas the internal diffu- Biochem. Microbiol. 2006, 42, 145–149.
sion limitations could be minimized by decreasing the 11. Kovalenko, G. A.; Perminova, L. V.; Terentyeva, T. G.;
size of the support granules below 1 mm. Under these Plaksin, G. V. Appl. Biochem. Microbiol. 2007, 43, 374–
conditions the maximum productivity of the IVR 378.
exceeded that of the packed-bed reactor by a factor of 12. Russian Patent 1706690 RU. Publ. 23.01, 1992.
13. Russian Patent 2064889 RU. Publ.11.02, 1993.
1.2–1.5. 14. Kuvshinov, G. G.; Mogilnykh, Yu. G.; Kuvshinov, D. G.;
A technological scheme for the production of starch Yermakov, D. Yu.; Yermakova, M. A.; Salanov, A. N.;
treacle with the dextrin hydrolysis stage carried out het- Rudina, N. A. Carbon 1999, 32, 1239–1246.
erogeneously using the ‘Glucoamylase on Sibunit’ bio- 15. Kovalenko, G. A.; Kuznetsova, E. V.; Mogilnykh, Yu. I.;
catalyst and an immersed vortex reactor has been Andreeva, I. S.; Kuvshinov, D. G.; Rudina, N. A. Carbon
2001, 39, 1133–1143.
constructed and tested on a laboratory scale. The advan- 16. Kovalenko, G. A.; Kuvshinov, D. G.; Komova, O. V.;
tages of the new technological scheme are significantly Simakov, A. V.; Rudina, N. A. In Nanoengineered
accelerated dextrin hydrolysis, energy and resource sav- Nanofibrous Materials; Güceri, S. I., Gogotsi, I. U. G.,
ing and high quality of the final products due to the lack Kuznetsov, V. V., Eds.; Klüver Academic: Dordrecht,
of protein impurities and easily regulated carbohydrate 2004; pp 169, 251–256.
17. Kovalenko, G. A.; Sukhinin, S. V.; Perminova, L. V.
composition of the treacle. Industrial Application of Biotechnology; Nova Science:
New York, 2006, pp 45–53.
18. Russian Patent 53670 RU. Publ. 27.05. 2006.
Acknowledgments 19. Bondareva, G. V.; Gulko, N. V. J. Appl. Chem. 1997, 70,
953–956.
20. Dandekar, A.; Baker, R. T. K.; Vannice, M. A. Carbon
The authors are grateful to Professor George V. Plaksin 1998, 36, 1821–1831.
and Professor Gennady G. Kuvshinov who donated