Kinetics of glycerol biotransformation to dihydroxyacetone by immobilized Gluconobacter oxydans and effect of reaction conditions

Article abstract of DOI:10.1135/cccc20071269

Biotransformation of glycerol to 1,3-dihydroxyacetone was carried out in an isothermal isochoric batch reactor with Gluconobacter oxydans immobilized in poly(vinyl alcohol) gel capsules. The reaction course was described with a three-step kinetic model. Two reaction schemes were proposed and compared with 8 kinetic experiments at 25°C. The experimental dependences of glycerol and dihydroxyacetone concentrations on reaction time were simulated very well by the autocatalytic model. The effects of reaction temperature and initial concentrations of yeast extract and glycerol were studied. Temperature 25-30°C, initial yeast extract concentration 2-4 g l^-1 and initial glycerol concentration 20-50 g l^-1 were found as optimal. The determined rate constants can be used to advantage for industrial production of dihydroxyacetone from glycerol.

Full text of DOI:10.1135/cccc20071269

Kinetics of Glycerol Biotransformation
1269
KINETICS OF GLYCEROL BIOTRANSFORMATION TO
DIHYDROXYACETONE BY IMMOBILIZED Glu con oba cter
oxyda n s AND EFFECT OF REACTION CONDITIONS
1,
Jiří RAŠKA, František SKOPAL, Karel KOMERS * and Jaroslav MACHEK
Department of Physical Chemistry, Faculty of Chemical Technology, University of Pardubice,
1
nám. Čs. legií 565, 532 10 Pardubice, Czech Republic; e-mail: karel.komers@upce.cz
Received February 27, 2007
Accepted April 3, 2007
Biotran sform ation of glycerol to 1,3-dih ydroxyaceton e was carried out in an isoth erm al
isoch oric batch reactor with Gluconobacter oxydans im m obilized in poly(vin yl alcoh ol) gel
capsules. Th e reaction course was described with a th ree-step kin etic m odel. Two reaction
sch em es were proposed an d com pared with 8 kin etic experim en ts at 25 °C. Th e experim en -
tal depen den ces of glycerol an d dih ydroxyaceton e con cen tration s on reaction tim e were
sim ulated very well by th e autocatalytic m odel. Th e effects of reaction tem perature an d in i-
tial con cen tration s of yeast extract an d glycerol were studied. Tem perature 25–30 °C, in itial
yeast extract con cen tration 2–4 g l–1 an d in itial glycerol con cen tration 20–50 g l–1 were
foun d as optim al. Th e determ in ed rate con stan ts can be used to advan tage for in dustrial
production of dih ydroxyaceton e from glycerol.
Keyw ord s: Dih ydroxyaceton e; Glycerol; Biotran sform ation s; Kin etics; Im m obilized m icro-
organ ism s; Biotech n ology.
Today m icrobial an d ch em ical tech n ologies com pete. Som e products can -
n ot be syn th etically produced econ om ically en ough an d som e ch em ical
procedures require m an y security precaution s. In th ese cases th e m icrobial
m eth ods are preferred due to th eir h igh ly specific course an d sligh t in flu-
en ce on en viron m en t. Th is h olds also for th e syth esis of 1,3-dih ydroxy-
aceton e (D) from glycerol (G)1.
Th e strictly aerobic m icroorgan ism Gluconobacter oxydans (G. oxydans),
wh ich is able to deh ydrogen ate various types of m on osacch arides, alcoh olic
sugars, aliph atic an d cyclic alcoh ols, is used for in dustrial production of D
from G1–12. D is presen tly produced m ostly by biotran sform ation (biological
oxidation ) of G with free cells of G. oxydans. D is used as a com pon en t of
bron zin g lotion s an d as a raw m aterial for th e production of ph arm aceuti-
cals. In th e in dustrial ferm en tation th e fed-batch reactors are often used,
wh ich can reduce th e effect of th e substrate in h ibition on th e process pro-
Collect. Czech. Chem. Commun. 2007, Vol. 72, No. 9, pp. 1269–1283
© 2007 Institute of Organic Chemistry and Biochemistry
doi:10.1135/cccc20071269

1270
Raška, Skopal, Komers, Machek:
ductivity1. Th e m ost m odern biotech n ologies utilize th e im m obilisation of
m icrobial cultures. Its advan tages are an in crease in biocatalyst con cen tra-
tion an d im provem en t of tech n ological properties. Th ere are m an y im m o-
bilisation tech n iques. Th e m ost used on es are im m obilisation in a biofilm
on th e surface of a suitable carrier an d en casem en t in to sem iperm eable cap-
sules (en capsulation ), e.g. by th e Len tiCats tech n ology13. Th e last tech -
n ique, usin g th e poly(vin yl alcoh ol) capsules, was used in th is work, wh ich
deals with
– th e form al kin etics of th e biotran sform ation of G to D by en capsulated
G. oxydans in th e isoth erm al isoch oric batch reactor. Th is description is
based on experim en tal depen den ces of G an d D m ass con cen tration s (G
an d D, both in g l–1) on reaction tim e t, an d
– th e in fluen ce of in itial con cen tration s of substrate an d yeast extract an d
tem perature of th e reaction m ixture.
THEORY
Because th e reaction m ixture durin g th e G biotran sform ation to D is h eter-
ogen eous, th e total tim e course m ust be a superposition of con secutive
steps, such as diffusion an d adsorption of G an d D on th e Len tiCats cap-
sules (L), an d bioch em ical tran sform ation of G to D. It is n ecessary to fin d
th e rate-determ in in g (slowest) step. Th erefore, th e rate of G an d D diffusion
th rough th e L poly(vin yl alcoh ol) m em bran e (PVA) in to th e capsules an d
back in to th e reactor volum e was prelim in ary determ in ed. It was foun d
th at both processes are m uch faster th an th e bioch em ical tran sform ation
proper. Th erefore it can be assum ed th at th e slowest step in th e wh ole pro-
cess is th e ch em ical reaction 14. If th e biotran sform ation proceeds well,
glyceric acid (A) is also form ed7 besides D. Th erefore th e reaction m ixture
m ust be m ain tain ed at con stan t pH value with an alkali. Glyceric acid is
form ed im m ediately from th e reaction start, wh en n o D is yet produced. It
can be assum ed th at A is form ed in a side (parallel) reaction besides th e oxi-
dation of G to D.
An oth er side step is th e con sum ption of G by th e in tern al m etabolism of
th e m icroorgan ism in cytoplasm . G serves as th e structural m aterial an d en -
ergy source for G. oxydans. In th e followin g m odels, th e product of th is m e-
tabolism is m arked as M. It was verified th at for th e sim plest reaction
m odel th e processes producin g A an d M can be approxim ately described by
th e first-order reaction with respect to G 10,14
.
All our experim en tal depen den ces G an d D vs t obtain ed in th e batch re-
actor are of sigm oidal form (Figs 2, 3, 5, 7). Th erefore, both kin etic m odels
Collect. Czech. Chem. Commun. 2007, Vol. 72, No. 9, pp. 1269–1283

Kinetics of Glycerol Biotransformation
1271
m en tion ed below were first tested for th e existen ce of th e in flexion poin t
on th e th eoretical depen den ces G an d D vs t.
As th e first reaction sch em e was proposed th e sim plest m odel optim ally
expressin g th e ph ysical prin ciple of th e reaction .
Catalytic Model
Th is m odel presum es th ree side steps.
O2, k1
O2, k2
O2, k3
G + E → D + E (1)
G → M + E (2)
G → A + E (3)
In th e first step (1) G reacts with th e en zym es E in cluded in en capsulated
G. oxydans un der D production . Th e secon d step (2) expresses form ally th e
growth of G. oxydans an d th e gain of en ergy at th e expen se of G. In th e
th ird step (3), th e experim en tally detected glyceric acid A is gen erated. By
rath er com plicated m ath em atical an alysis of th e differen tial kin etic equa-
tion s describin g th is m odel,
–dG/dt = G(k1E + k2 + k3)
dE/dt = (k2 + k3)G
dA/dt = k3GE ,
it could be dem on strated, th at th e depen den ces of G an d D vs t h ave n o
poin t of in flexion an d th us can n ot sim ulate th e obtain ed experim en ts.
Th erefore, th is m odel is un suitable.
Th e secon d, form al, purely statistical reaction sch em e is th e autocatalytic
m odel, treated in th e n ext paragraph .
Autocatalytic Model
In th is m odel G reacts form ally with E un der production of n ot on ly A, D
an d M but also regen eration of E.
O2, kD
O2, kM
O2, kA
G + E → 2 E + D (4)
G + E → 2 E + M (5)
G + E → 2 E + A (6)
Let G, E, D, M be th e actual m ass con cen tration s of G, E, D, M (in g l–1),
g, e in itial m ass con cen tration s of G, E, an d I stan ds for actual con cen tra-
Collect. Czech. Chem. Commun. 2007, Vol. 72, No. 9, pp. 1269–1283

1272
Raška, Skopal, Komers, Machek:
tion of an y given reagen t. Th en for G an d D followin g differen tial kin etic
equation s in th e batch reactor can be written
–dG/dt = (kD + kM + kA)GE = dE/dt
(7)
dD/dt = kDGE
(8)
wh ere th e rate con stan ts k, kD an d kM are in l g–1 s–1. After in troduction of
relative con cen tration s ϕi = I/g, we h ave th e kin etic relation s of th e con cen -
tration –con cen tration type
e
1 + = ϕG + ϕE
(9)
g
ϕG + ϕD + ϕM + ϕA = 1.
(10)
In tegration of th e differen tial kin etic equation s (7) an d (8) gives
e
1 +
g
ϕ G
=
(11)
e
1 + exp (kgt)
g
e
exp (kgt)
g
e
ϕ E = 1 +
(12)
(13)
e
g
1 + exp (kgt)
g
kD
exp (kgt) − 1
e
ϕ D
=
e
k
g
1 + exp (kgt)
g
wh ere k = kD + kM + kA.
Th ese relation s sh ow th e sigm oidal course with th e in flexion poin t. After
in troduction of param eters P1, P2 an d P3, wh ich are optim ized by n on lin ear
regression , Eqs (11) an d (13) acquire th e form s
1 + P
1
ϕ G
=
(14)
1 + P exp (tP2 )
1
Collect. Czech. Chem. Commun. 2007, Vol. 72, No. 9, pp. 1269–1283

Kinetics of Glycerol Biotransformation
1273
exp (tP2 ) − 1
ϕ D = P P3
(15)
1
1 + P exp (tP2 )
1
wh ere P1 = e/g < 1, P2 = kg an d P3 = kD/k.
For th e in flexion poin t, we m ay write
e
ϕ G
= 1 +
/ 2
(16)
in f
g
an d its value m ust be sm aller th an 1, wh ich is possible on ly wh en e/g < 1.
For t → ∞ it could be written
kD
k
e
ϕ G = 0 (17a)
ϕ E ∞ = 1 + > 1 (17b)
ϕ D ∞
=
< 1 (17c).
∞
g
Th is m ean s th at at th e en d of th e reaction , G is spen t com pletely, th e con -
cen tration of E in creases an d D is form ed in th e relative con cen tration
sm aller th an 1.
EXPERIMENTAL
Ch em icals
G. oxydans CCM 1783 (Czech collection of m icroorgan ism s, Brn o, Czech Republic) im m o-
bilized in Len tiCats capsules from poly(vin yl alcoh ol) (Len tiKat’s Biotech n ologies a.s.,
Prah a, Czech Republic), glycerol p.a., potassium h ydroxide p.a. 85% w/w, sulfuric acid p.a.
98% w/w, eth an ol den aturated (all Lach em a Brn o, Czech Republic), yeast extract Ferm Tech
(Merck, Darm stadt, Germ an y), defoam er An tispum in WA (Radka Gm bH, Pardubice, Czech
Republic), Persteril (32–36% peracetic acid, 5–12% h ydrogen peroxide, 25% acetic acid, <1%
sulfuric acid, Peroxides, Sokolov, Czech Republic).
Apparatus an d Meth ods
All ferm en tative reaction s were carried out in th e apparatus depicted in Fig. 1. Every startin g
solution used for biotran sform ation in cluded a m ixture of th ree com pon en ts: G, yeast ex-
tract (YE) an d water. Th e defoam er An tispum in -WA was added to all startin g solution s to re-
m ove th e biological foam .
Un der sterile con dition s15, 500 m l of aqueous solution con tain in g 58.4 (±0.6) g l–1 G an d
2.1 (±0.2) g l–1 YE were in troduced in to th e reactor at a ch osen tem perature. Th e pH was ad-
justed an d th en h eld at 5.1. Th e flow rate of air was 10 l h –1 with th e overpressure ca. 25 Torr.
Th e reaction m ixture was agitated with a propeller at 400 rpm . Th e biotran sform ation was
started by th e addition of a suitable am oun t of settled en capsulates with im m obilized
G. oxydans (see below). Th e pH decrease due to production of A was con tin ually com pen -
sated by addition of aqueous KOH. In th e liquid sam ples (ca. 5 m l) taken from th e reaction
Collect. Czech. Chem. Commun. 2007, Vol. 72, No. 9, pp. 1269–1283

1274
Raška, Skopal, Komers, Machek:
m ixture in suitable tim e in tervals, th e reaction was stopped by acidification with 0.1 M HCl
to pH 2 an d th e sm all am oun t of solid G. oxydans released from th e pellets was rem oved by
cen trifugation . Th e con cen tration s of G an d D in th e sam ples were determ in ed sim ulta-
n eously by HPLC isocratic an alysis.
An alytical Meth ods
HPLC isocratic determination of glycerol and dihydroxyacetone. HA pum p LCP 4000 (Ecom
Prah a, Czech Republic), refractom etric detector RIDK 102 (Laboratorn í přístroje Prah a, Czech
Republic), ion ex colum n Watrex IEX in H-form 250 × 8 m m , m obile ph ase 0.01 M H2SO4,
flow rate 0.3 m l m in –1, sam ple dose 3 µl. Th e reten tion tim es of m axim a of G an d D peaks
were 23 an d 23.9 m in , respectively. Th e con cen tration s of G an d D were determ in ed usin g
calibration curves – area vs con cen tration with h elp of th e PC program CSW 32.
Estimation of biomass content in capsules. Capsules with im m obilized G. oxydans were th in
len tils ca. 5 m m in diam eter. A suspen sion of th e capsules in aqueous 50% (w/w) solution of
G (at th is con cen tration th e bioreaction is in h ibited) was kept at 5 °C. A suitable am oun t of
capsules was wash ed with water. Th e water rem ain in g on th e capsules was partly rem oved
by dryin g between sh eets of filter paper. Com plete dryin g of capsules is n ot appropriate due
to irreversible ch an ges in th e PVA structure. PVA coatin g of 2–5 g of partly dried capsules
(m oist capsules) was dissolved in 40 m l of h ot water, th e suspen sion of free G. oxydans was
cen trifuged in tubes (2.5 × 10 cm at 6000 rpm ) an d wash ed th ree tim es with 30 m l of water.
Th e sedim en t was dried in a Petri dish at 105 °C for 3 h . From th e m ass of th e dry sedim en t
an d in itial m ass of m oist capsules, con cen tration of m icroorgan ism s (m g of dry biom ass
G. oxydans per 1 g of m oist capsules) was calculated. Th e ratio was used for calculation of
th e in itial am oun t of m oist capsules represen tin g th e ch osen in itial G. oxydans con cen tra-
tion l (in g l–1) in th e reaction m ixture. Tem perature, pH an d con cen tration of dissolved ox-
ygen in th e reaction m ixture were con tin uously m easured with a com bin ed sen sor.
FIG. 1
1 stock KOH aqueous solution, 2 pH-stat, 3 ferm entation batch reactor, 4 glass electrode, 5 oxygen
electrode, 6 dosin g of KOH solution , 7 glass tube with frit, 8 defoam er, 9 com pressor, 10
m ultim eter, 11 PC, 12 electric propeller, 13 reflux con den ser, 14 m an om eter, 15 m an ostat
with silicon e oil, 16 output from m an ostat
Collect. Czech. Chem. Commun. 2007, Vol. 72, No. 9, pp. 1269–1283

Kinetics of Glycerol Biotransformation
1275
Note: Because of un certain con ten t of water in m oist capsules an d low weigh t of dry m i-
croorgan ism s in th e dry G. oxydans sedim en t, th is m eth od gives rath er an estim ate th an th e
exact biom ass con ten t in capsules.
RESULTS AND DISCUSSION
Th e fit of experim en tal an d th eoretical depen den ces was ch ecked by th eir
correlation coefficien ts (n ever sm aller th an 0.988) an d visually (see Figs 2,
3, 5 an d 7 below). As quan titative param eters, optim um values of rate con -
stan ts k an d kD were obtain ed.
Kinetics of Biotransformation
Th e catalytic m odel is m ost sim ple an d correspon ds well with th e ph ysical
idea of th e reaction . However, th e th eoretical equation s con cen tration vs
tim e h ave n ot th e experim en tally foun d sigm oidal ch aracter an d can n ot
th erefore sim ulate th e experim en t.
Verification of th e autocatalytic m odel was based on eigh t experim en ts
perform ed in th e batch reactor described in Fig. 1. Th e obtain ed experim en -
tal tim e depen den ces of ϕG an d ϕD were evaluated accordin g to Eqs (14)
an d (15) by n on lin ear regression with software Origin 7.0. Th e total kin etic
results obtain ed from th ese experim en ts are dem on strated in Table I. An
TABLE I
2
2
Kin etic results for th e autocatalytic m odel. RG an d RDHA are th e determ in ation coefficien ts.
Th eir best value is 1
k × 106
kD × 107
2
2
Experim en t
e/g × 103
RG
RDHA
l g–1 s–1
EB1
EB2
EB3
EB4
EB5
EB6
EB7
EB8
3.6
1.32
1.18
1.29
1.04
1.26
1.13
1.57
1.10
7.54
6.57
6.73
6.19
7.17
7.06
8.75
6.47
0.99878
0.99987
0.99989
0.99977
0.99995
0.99991
0.99990
0.99989
0.99844
0.99928
0.99528
0.99952
0.99772
0.99868
0.99954
0.99902
1.94
2.84
3.97
4.20
4.71
3.78
4.64
Collect. Czech. Chem. Commun. 2007, Vol. 72, No. 9, pp. 1269–1283

1276
Raška, Skopal, Komers, Machek:
exam ple of con sisten cy of depen den ces from experim en t EB1 an d th eir
sim ulation by th eoretical relation s (14) an d (15) is sh own in Fig. 2.
Th e values of average determ in ation coefficien ts in Table I, RG = 0.9997
an d RD = 0.9984 (i.e. correlation coefficien ts 0.9998 an d 0.9992), an d also
th e graph ical agreem en t of experim en tal an d th eoretical data in all 8 exper-
im en ts con firm th e validity of th e purely statistical autocatalytic m odel.
Th is m odel in cludes (see Eq. (17)) also th e growth of m icroorgan ism s dur-
in g th e reaction . Th is growth is con firm ed by th e in creasin g in itial ratios
e/g in Table I in experim en ts EB1–EB8, wh ich were perform ed with th e
sam e am oun t of poly(vin yl alcoh ol) pellets in cludin g in creasin g am oun t of
G. oxydans.
Effect of Reaction Conditions
In fluen ce of In itial Yeast Extract Con cen tration
Th e tim e depen den ces of G an d D for four in itial con cen tration s of steril-
ized aqueous YE ([YE] = 1, 2.5, 4 an d 5 g l–1) were m easured. Tem perature
25 °C, in itial con cen tration s of G (g = 58.6–58.8 g l–1) an d of biom ass (l =
3.78 g l–1) were kept con stan t for each experim en t. Th e obtain ed curves ϕG
an d ϕD vs t are presen ted in Fig. 3. Param eters P2 an d P3 were foun d by n on -
lin ear regression an d from th ese param eters th e values of k an d kD were
FIG. 2
Autocatalytic m odel, experim en t EB1. Sim ulation (full lin es) of experim en tal depen den ces
(poin ts) ϕG (᭿) an d ϕD (ꢀ) vs tim e (t) accordin g to Eqs (14) an d (15)
Collect. Czech. Chem. Commun. 2007, Vol. 72, No. 9, pp. 1269–1283

Kinetics of Glycerol Biotransformation
1277
calculated (Table II). Th e depen den ce of rate con stan ts k an d kD on con cen -
tration of YE is dem on strated in Fig. 4.
With in creasin g YE con cen tration (up to ca. 3.5 g l–1), th e rates of
biotran sform ation reaction (k) an d D production (kD) in crease. For h igh er
YE con cen tration s, both rate con stan ts sh arply decrease. Th e presen ce of
YE in th e bioreaction is n ecessary. In th e absen ce of n utrien ts from YE, n o
en zym e respon sible for biotran sform ation of G to D is produced. On th e
con trary, at too h igh YE con cen tration s, th e rate of D production rapidly
falls. Th is can be caused, e.g., by th e reduced perm eability of capsules to
reaction com pon en ts. In th e sam e biocon version with free cells, th e m axi-
TABLE II
Effect of in itial con cen tration of YE on kin etic param eters k, kD
[YE]0, g l–1
Param eter
l g–1 h –1
1
2.5
4
5
k
0.00462
0.00323
0.005343
0.003572
0.005339
0.00373
0.00386
kD
0.002587
FIG. 3
Depen den ces of relative m ass con cen tration of G an d DHA (ϕG an d ϕD) vs tim e (t) for four in i-
tial con cen tration s of yeast extract [YE]0. Poin ts, experim en t; lin es, sim ulation s accordin g to
Eqs (14) an d (15)
Collect. Czech. Chem. Commun. 2007, Vol. 72, No. 9, pp. 1269–1283

1278
Raška, Skopal, Komers, Machek:
m um [YE]0 = 5 g l–1 used in th is work was dosed routin ely with out an y in h i-
bition effect.
In fluen ce of In itial Con cen tration of Glycerol
Th e tim e depen den ces of G an d D for four in itial con cen tration of glycerol
(g = 20.6, 58.6, 110.8 an d 219.8 g l–1) were m easured. Tem perature 25 °C,
in itial con cen tration of YE (2.5 g l–1) an d of biom ass (l = 4.3 g l–1) were kept
con stan t in each experim en t. Th e resultin g curves of ϕG an d ϕD vs t are
sh own in Fig. 5. Param eters k an d kD are presen ted in Table III. Th e depen d-
en ce of rate con stan ts on in itial con cen tration of glycerol is given in Fig. 6.
TABLE III
Effect of in itial con cen tration of G on kin etic param eters k, kD
g, g l–1
Param eter
l g–1 h –1
20.6
58.6
110.8
219.8
k
0.01436
0.00862
0.00510
0.00114
0.00064
0.00224
0.00075
kD
0.003421
1
2
FIG. 4
Effect of in itial yeast extract con cen tration , [YE]0, on th e rate con stan ts k (1) an d kD (2).
Poin ts, experim en t; see Table II
Collect. Czech. Chem. Commun. 2007, Vol. 72, No. 9, pp. 1269–1283

Kinetics of Glycerol Biotransformation
1279
FIG. 5
Depen den ces of relative con cen tration of G an d DHA (ϕG an d ϕD) vs tim e (t) for four in i-
tial con cen tration s of glycerol g. Poin ts, experim en t; lin es, sim ulation s accordin g to Eqs (14)
an d (15)
1
2
FIG. 6
Effect of in itial glycerol con cen tration g on th e rate con stan ts k (1) an d kD (2). Poin ts, experi-
m en t; see Table III
Collect. Czech. Chem. Commun. 2007, Vol. 72, No. 9, pp. 1269–1283

1280
Raška, Skopal, Komers, Machek:
Th e rate of D production decreases with in creasin g in itial con cen tration
of G. At in itial con cen tration g = 20.6 g l–1, th e reaction was com plete (G =
0) after 12 h . At g = 219.8 g l–1 th is tim e period was fifteen tim es lon ger. As
we can see from Figs 5 an d 6, at g > 100 g l–1 th e reaction tim e in creased
an d th e kD value m arkedly decreased. On th e oth er h an d, a very h igh value
of g in creased th e overall biotran sform ation rate, givin g probably oth er
products th an D. Th e in itial con cen tration of G over 100 g l–1 is th erefore
disadvan tageous.
In fluen ce of Reaction Tem perature
Th e tim e depen den cies of G an d D for th ree reaction tem peratures (20, 25
an d 30 °C) were m easured. Th e in itial con cen tration of YE (2.5 g l–1), in itial
con cen tration s of biom ass (l = 4.12 g l–1) an d of glycerol (g = 58.6 g l–1) were
kept con stan t in each experim en t. Th e obtain ed curves ϕG an d ϕD vs t are
presen ted in Fig. 7. Param eters k an d kD are given in Table IV. Th e depen d-
en ces of th e rate con stan ts on reaction tem perature are presen ted in Fig. 8.
Th e depen den ces of biotran sform ation rate con stan ts on tem perature in
Fig. 6 sh ow th at th e optim al reaction tem perature lies between 25 an d 30 °C.
In th is in terval th e differen ce in th e rate con stan ts is sm all. Th e lower solu-
FIG. 7
Depen den ces of relative con cen tration of G an d DHA (ϕG an d ϕD) vs tim e (t) at th ree various
tem peratures. Poin ts, experim en t; lin es, sim ulation s accordin g to Eqs (14) an d (15)
Collect. Czech. Chem. Commun. 2007, Vol. 72, No. 9, pp. 1269–1283

Kinetics of Glycerol Biotransformation
1281
bility of oxygen in th e reaction m ixture at 30 °C h ad to be com pen sated by
m ore in ten sive aeration . Tem peratures over 30 °C could n ot be studied due
to th e solubility of PVA capsules in water. At 15 °C, th e biotran sform ation
practically did n ot even start after 10 h , i.e. n o perceptible decrease in G
an d in crease in D could be m easured.
TABLE IV
Effect of tem perature on kin etic param eters k, kD
Tem perature, °C
Param eter
l g–1 h –1
20
25
30
k
0.00311
0.00182
0.00303
0.00206
0.00334
0.00220
kD
1
2
FIG. 8
Effect of tem perature on th e rate con stan ts k (1) an d kD (2). Poin ts, experim en t; see Table IV
Collect. Czech. Chem. Commun. 2007, Vol. 72, No. 9, pp. 1269–1283

1282
Conclusions
Raška, Skopal, Komers, Machek:
Kin etics of glycerol biotran sform ation s in to 1,3-dih ydroxyaceton e by G.
oxydans im m obilized in poly(vin yl alcoh ol) pellets was at 25 °C successfully
described by th e form al autocatalytic m odel con sistin g of th ree parallel
steps wh ich produce, besides dih ydroxyaceton es, also glyceric acid an d con -
sum e a part of glycerol for th e growth of th e m en tion ed m icroorgan ism .
Th e rate con stan t of dih ydroxyaceton e production step kD an d th e sum of
rate con stan ts k of all th ree steps of th is m odel were calculated. Th e deter-
m in ed values of rate con stan ts can be used to calculate th e optim al con cen -
tration s of G an d D in th e production of 1,3-dih ydroxyaceton e by G. oxydans
en capsulated in poly(vin yl alcoh ol) gel.
Th e effect of in itial yeast extract an d glycerol con cen tration s an d reaction
tem perature on th e course of th is biotran sform ation was studied. Th e pro-
posed th ree-step kin etic m odel of th e biotran sform ation was experim en -
tally verified also for th e tem perature in terval 20–30 °C. In creasin g in itial
con cen tration of yeast extract (up to 3.5 g l–1) in creases an d of glycerol
m arkedly decreases th e rate of production of 1,3-dih ydroxyaceton e. At tem -
peratures ≤15 °C, n o 1,3-dih ydroxyaceton e is produced, tem peratures over
30 °C already dissolve th e PVA capsules in water. In th e ran ge of 25–30 °C,
th e rate of th e 1,3-dih ydroxyaceton e production is practically con stan t.
Optim al con dition s of th e studied biocon version are: reaction tem pera-
ture 25–30 °C, in itial yeast con cen tration 2–4 g l–1 an d in itial con cen tration
of glycerol 20–50 g l–1.
SYMBOLS
A
D, D
E
G, G
g, e
ki
glyceric acid
1,3-dih ydroxyaceton e, m ass con cen tration of D (g l–1
en zym e(s) in G. oxydans
)
glycerol, m ass con cen tration of G (g l–1
)
in itial m ass con cen tration s of G, E (g l–1
rate con stan t of i-th reaction (l g–1 h –1
)
)
k = kD + kM + kA
[L]0
in itial m olar con cen tration of G. oxydans (m ol l–1
)
l
M
in itial m ass con cen tration s of dry G. oxydans (g l–1
products of m etabolic processes
poly(vin yl alcoh ol)
)
PVA
YE, [YE]0
ϕG, ϕD
yeast extract, in itial m ass con cen tration of YE (g l–1
)
relative (dim en sion less) con cen tration s G/g, D/g
Collect. Czech. Chem. Commun. 2007, Vol. 72, No. 9, pp. 1269–1283

Kinetics of Glycerol Biotransformation
1283
This work was financially supported by the Ministry of Education, Youth and Sports of the Czech
Republic as a part of the research project MSMOO 21627502.
REFERENCES
1. Hekmat D., Bauer R., Fricke J.: Bioprocess Biosyst. Eng. 2003, 26, 109.
2. Bories A., Claret C., Soucaille P.: Process Biochem. 1991, 26, 243.
3. Claret C., Bories A., Soucaille P.: Curr. Microbiol. 1992, 25, 149.
4. Claret C., Salmon J. M., Romieu C., Bories A.: Appl. Microbiol. Biotechnol. 1994, 41, 359.
5. Švitel J., Šturdík E.: J. Ferment. Bioeng. 1994, 78, 351.
6. Ohrem H. L., Voss H.: Process Biochem. 1996, 31, 295.
7. Sommer K., Bohlmann U., Bohnet M.: Chem. Eng. Technol. 1998, 21, 144.
8. Smrž M., Kabát J., Marek A., Šimoník M.: Czech. 290 334 (2002); Chem. Abstr. 2004, 140,
58549.
9. Chotani G.: Biochim. Biophys. Acta 2000, 1543, 434.
10. Gupta A., Singh. V. K., Quazi G. N., Kumar. A.: Process Biochem. 2001, 3, 445.
11. Navrátil M., Tkáč J., Švitel J., Danielsson B., Šturdík E.: Process Biochem. 2001, 36, 1045.
12. Tkáč J., Navrátil M., Šturdík E., Gemeiner P.: Enzyme Microb. Technol. 2001, 28, 383.
13. Stloukal R. (LentiKat’s Biotechnologies a.s., Prague): Private communication.
14. Raška J.: Ph.D. Thesis. Faculty of Chemical Technology, University of Pardubice,
Pardubice 2005.
15. Raška J., Skopal F., Komers K., Machek J., Stloukal R.: Sci. Pap. Univ. Pardubice A 2005,
11, 145.
Collect. Czech. Chem. Commun. 2007, Vol. 72, No. 9, pp. 1269–1283