Appl Biochem Biotechnol (2014) 172:2769–2785

DOI 10.1007/s12010-014-0723-7

A Study of the Effects of Aeration and Agitation

on the Properties and Production of Xanthan Gum
from Crude Glycerin Derived from Biodiesel Using
the Response Surface Methodology

Denilson de Jesus Assis & Líllian Vasconcelos Brandão &

Larissa Alves de Sousa Costa & Tamiris Vilas Boas Figueiredo &
Luciane Santos Sousa & Francine Ferreira Padilha &
Janice Izabel Druzian

Received: 30 September 2013 / Accepted: 2 January 2014 /

Published online: 17 January 2014
# Springer Science+Business Media New York 2014

Abstract The effects of aeration and agitation on the properties and production of xanthan
gum from crude glycerin biodiesel (CGB) by Xanthomonas campestris mangiferaeindicae
2103 were investigated and optimized using a response surface methodology. The xanthan
gum was produced from CGB in a bioreactor at 28 °C for 120 h. Optimization procedures
indicated that 0.97 vvm at 497.76 rpm resulted in a xanthan gum production of 5.59 g L−1 and
1.05 vvm at 484.75 rpm maximized the biomass to 3.26 g L−1. Moreover, the combination of
1.05 vvm at 499.40 rpm maximized the viscosity of xanthan at 0.5 % (m/v), 25 °C, and 25 s−1
(255.40 mPa s). The other responses did not generate predictive models. Low agitation
contributed to the increase of xanthan gum production, biomass, viscosity, molecular mass,
and the pyruvic acid concentration. Increases in the agitation contributed to the formation of
xanthan gum with high mannose concentration. Decreases in the aeration contributed to the
xanthan gum production and the formation of biopolymer with high mannose and glucose
concentrations. Increases in aeration contributed to increased biomass, viscosity, and formation
of xanthan gum with greater resistance to thermal degradation. Overall, aeration and agitation
of CGB fermentation significantly influenced the production of xanthan gum and its
properties.

D. de Jesus Assis (*) : L. V. Brandão : L. A. de Sousa Costa

Department of Chemical Engineering, Polytechnic School, Federal University of Bahia,
Aristides Novis Street, n° 2, Second Floor, Federação, Salvador, Bahia 40210-630, Brazil
e-mail: denilsonengal@gmail.com

T. V. B. Figueiredo : L. S. Sousa : J. I. Druzian

Department of Bromatological Analysis, College of Pharmacy, Federal University of Bahia,
Barão of Geremoabo Street, s/n, Ondina, Salvador, Bahia 40171-970, Brazil

F. F. Padilha
Technology and Research Institute, Tiradentes University, Murilo Dantas Street, n° 300,
Aracajú, Sergipe 49040-020, Brazil
2770 Appl Biochem Biotechnol (2014) 172:2769–2785

Keywords Xanthan gum . Crude glycerin . Aeration . Agitation . Properties

Introduction

Xanthan gum is an extracellular biopolymer produced by various gram-negative bacteria of the

genus Xanthomonas [1]. Due to its structure, xanthan gum exhibits pseudoplastic properties:
high viscosity and solubility, enhanced stability over a wide range of temperature and pH
values, and compatibility with many salts and other polysaccharides. This biopolymer is a
favorable thickener and stabilizer of emulsions and suspensions used in the food industry,
pharmaceutical formulations, paper milling, ceramic glazes, and agricultural products [2, 3].
Recently, the global xanthan gum market has been progressively increasing at an annual rate of
5–10 % [4].
The main chain of xanthan gum is composed of D-glucose (β-1,4) units that form a
cellulose backbone with trisaccharide side chains composed of mannose (β-1,4) and glucu-
ronic acid (β-1,2); mannose is attached to alternate glucose residues in the backbone by α-1,3
linkages [2, 5].
The composition structure, molecular mass, and viscosity depend on several factors, such as
the composition of the culture medium, carbon and nitrogen sources, mineral salts, trace
elements, type of Xanthomonas strains, and fermentation conditions (time, pH, temperature,
oxygen concentration, and agitation rate) [6, 7].
Large-scale batch production processes of xanthan often use glucose as a substrate. Such
processes are highly productive [8], but the increasing market price and demand suggest that
glucose may no longer be an economically feasible raw material. Because the use of glucose or
sucrose significantly impacts the production costs of this polysaccharide [3], other sources
have also been tested as raw materials, such as raw cassava starch [9], date palm juice [4],
coconut juice, unmodified starch [2], potato residue [10], apple residue [11], acid whey [12],
and cassava serum [13].
Glycerin is a major by-product of the biodiesel manufacturing process. Generally, approx-
imately 4.53 kg of crude glycerin is created for every 45.3 kg of biodiesel produced [14]. The
worldwide biodiesel market is estimated to reach 37 billion gallons by 2016, which implies
that approximately 4 billion gallons of crude glycerol will be produced [15]. Thus, new uses
for crude glycerin are urgently needed.
The glycerin derived from biodiesel production is impure, and the significant cost of
purification prevents its use in the food and pharmaceutical industries [16]. A promising
alternative use of this by-product is the microbial conversion of crude glycerin into value-
added products through biotechnological processes, such as polyhydroxybutyrate polymers
[17], clavulanic acid [18], recombinant human erythropoietin [17], citric acid [19], and xanthan
gum [20].
The response surface methodology (RSM) technique can optimize complex process by
facilitating the arrangement and interpretation of experiments more easily compared to other
traditional methods [21].
This study reports the effects of aeration and agitation speed on the production and
properties of the xanthan gum production by Xanthomonas campestris from crude glycerin
derived from biodiesel (CGB) during batch fermentation. Accordingly, this research aimed to
optimize the selection process variables to maximize the dependent variables using response
surface methodology. Specifically, the individual and simultaneous effects of the aeration and
speed agitations were investigated on the production, apparent viscosity, composition, molec-
ular mass, and initial temperature degradation of the polymers.
Appl Biochem Biotechnol (2014) 172:2769–2785 2771

Materials and Methods

Microorganism and Inoculum Preparation

The Tropical Collection Culture of the Biological Institute (Campinas, São Paulo, Brazil)
supplied the X. campestris mangiferaeindicae 2103 strain isolated from Brazil and used in this
study. The strain was grown and maintained in yeast extract malt agar (YMA) with the
following composition (in grams per liter): 10.0 glucose, 3.0 yeast extract, 3.0 malt extract,
5.0 peptone, and 15.0 agar. During the preparation, the media pH was adjusted to 7.0 and then
was sterilized (121 °C/20 min). After growing for 48 h at 28 °C, the culture was maintained at
4 °C. To minimize losses in productivity and fluctuations in the results during subsequent
cultivations due to intra-population variability, the cell cultures were preserved in glycerol
(20 % v/v) at −80 °C. Each experiment was started by reviving one glycerol vial in vegetative
culture [22].
Inoculum cultures were prepared in yeast malt (YM) broth containing (w/v) 1.0 %
glucose, 0.5 % peptone, 0.3 % yeast extract, and 0.3 % malt extract [3]. The medium
pH was adjusted to 7.0 before autoclaving (121 °C/15 min). X. campestris
mangiferaeindicae 2103 cells were incubated in 250-mL Erlenmeyer flasks containing
50 mL of YM broth at 28±2 °C for 24 h. The flasks were placed in an orbital shaker
(Tecnal mod. TE-424, São Paulo, Brazil) at 180 rpm. Cell growth was monitored
spectrophotometrically (PerkinElmer model Lambda 20) by measuring the optical
density at 620 nm after 48 h of incubation until the cell concentration reached
1011 UFC mL−1.

Crude Glycerin Biodiesel Composition

Crude glycerin biodiesel (CGB) was supplied by the Biodiesel Pilot Plant of the State
University of Santa Cruz (Ilhéus, Bahia, Brazil). The following analyses were carried out on
the crude glycerin in triplicate: acidity (pH), volatile content at 105 °C, protein (Kjeldahl
method), ash [23], and total lipid content [24]. The carbohydrate content (glycerol) was
calculated by difference [100−(ash+protein+volatiles+lipid) percentages].

Xanthan Gum Production Medium and Cultivation Conditions

To enhance the process variables, all fermentations were conducted in a 4.5-L bioreactor
(Tecnal Mod. TecBio, Piracicaba, São Paulo, Brazil) containing 3.0 L of production medium
consisting of 2.0 % (v/v) CGB, 0.01 % (w/v) (NH2)2CO, 0.1 % (w/v) KH2PO4, and 0.1 % (v/v)
antifoam. The medium’s pH was adjusted to 7.0, and the medium was then sterilized. The
fermentation medium was inoculated (20 % v/v) at 28±2 °C, and the fermentation was allowed
to proceed for 120 h. Cell growth was estimated by measuring the absorbance of the
cell suspensions at 620 nm in a PerkinElmer model Lambda 35 spectrophotometer
(Norwalk, USA).

Biomass Concentration

The cells were collected after centrifugation at 18,800×g for 30 min (HITACHI, CR-GIII).
After the supernatant was discarded, the biomass was washed with 0.85 % NaCl solution and
re-centrifuged. This process was repeated twice. Finally, the cells were dried in an oven for
24 h and weighed.
2772 Appl Biochem Biotechnol (2014) 172:2769–2785

Xanthan Gum Recovery and Purification

The xanthan gum product was recovered from the supernatants of centrifugation by precipi-
tation with 98 % ethanol at a 3:1 ratio (v/v). The precipitated xanthan gum was collected and
dried in an oven (Tecnal TE 394/2, São Paulo, Brazil) at 30±2 °C for 72 h. The production was
expressed as grams per liter of fermentation broth. The aqueous xanthan gum solutions were
dialyzed (cutoff 12,000 Da) against purified water at 100 rpm and 25 °C for 72 h in an orbital
shaker (Tecnal TE-424, São Paulo, Brazil). The solutions were lyophilized (LIOBRAS L101)
for 48 h and then stored in hermetical flasks.

Apparent Viscosity of the Xanthan Gum Solution

To study the rheological characteristics of the gums, 0.5 % (m/v) xanthan gum solutions were
prepared using distilled water, stirred for 5 min to complete dissolution, and then maintained at
room temperature for 12 h before testing. The apparent viscosities were measured in a
concentric cylinder rheometer (Haake Rheotest mod 2.1, Medingen, Germany) coupled with
a wash bath for temperature control and shear rate of 25 to 1,000 s−1. The rheological data
were fitted to the Ostwald–de Waele model:

μ ¼ Kγ ðn−1Þ ð1Þ
where μ is the apparent viscosity, K is the consistency index, γ is the shear rate, and n is the
flow behavior index.

Experimental Methodology

The individual and interactive effects of aeration (X1) (0.5–1.5 vvm) and agitation (X2) (300–
700 rpm) on xanthan gum production (Y1), biomass (Y2), apparent viscosity (Y3), glucose
content (Y4), mannose content (Y5), glucuronic acid content (Y6), pyruvic acid content (Y7),
average molecular mass (Y8), and maximum thermal degradation temperature (Y9) were
determined using response surface methodology [25].
The procedure performed to determine the ideal agitation and aeration conditions to
optimize polymer production and properties followed a full factorial experimental design of
type 22, with three central points and four axial at a distance α=±1.412, totaling 11 experi-
ments. The levels (in coded values) were −1, 0, and +1, where 0 corresponded to the central
point. The coded values were calculated according to the following equation:

actual value−ðhigh level þ low levelÞ=2

Coded value ¼ ð2Þ
ðhigh level−low levelÞ=2

The responses of the dependent variables were analyzed using the Statistica software for
Windows version 7, and the significance level was 5 %. The levels chosen were based on
preliminary tests conducted by our research group.
Mathematical models were fitted to the experimental points and from the analysis of
variance (ANOVA) model. The predictive power of the regression was tested using the Fisher
(F) test and coefficients (R2) [25]. Fit testing and model predicting were set at a 5 %
significance level (P value ≤0.05).
To find the values that maximize or minimize each estimated response, the stationary point
of the surface was calculated by canonical analysis [25] to determine optimal combination of
aeration and agitation speed. Furthermore, the nature of the surface was evaluated based on the
Appl Biochem Biotechnol (2014) 172:2769–2785 2773

signals from the root characteristics of the quadratic equations (λ1e λ2) from the surface of the
second-order model.

Xanthan Gum Average Molecular Mass

The average molecular mass of xanthan gum was estimated by size-exclusion chromatography
(GPC HPLC systems (PerkinElmer Series 200; Shelton, USA)) with Shodex OHpak SB 803,
804, 805, and 806 columns in series (Kawasaki-ku, Japan) using 0.5 % (w/v) NaNO3 as the
eluant at a flow rate of 1 mL min−1. A Refractive Index (RI) PerkinElmer Series 200 (Shelton,
USA) was used as the detector. The column was calibrated with dextran patterns (102,000;
207,200; 431,800; 655,200; 759,400; 1,360,000; 2,025,000; 2,800,000; 3,450,000; and
5,900,000 Da) (American Polymer Patterns, USA). An 80-μL aliquot (0.3 % m/v) of dextran
pattern and xanthan gum aqueous solutions was injected in triplicate.
The molecular mass of xanthan gum was calculated using the calibration curve log
molecular mass (dextran patterns molecular mass)×retention time (RT) and compared with
the molecular mass of Sigma xanthan gum (reference).

Xanthan Gum Chemical Composition

The glucose, mannose, glucuronic acid, and pyruvate contents in xanthan gum synthesized by
X. campestris mangiferaeindicae 2103 were quantified. The samples (10.0 mg) were initially
hydrated with purified water (0.5 mL) for 12 h, hydrolyzed with 1 M trifluoroacetic acid (TFA)
at 0.5 mL for 10 h at 100 °C, subsequently dried with nitrogen gas, lyophilized to completely
remove any TFA residue, and dissolved with 1 mL chromatographic grade water.
To determine the sugar concentrations, the hydrolyzed polymer solutions were injected into
the HPLC-IR (PerkinElmer 200 series) using a Polypore Ca pre-column (30 mm×10 mm×
4.6 mm) followed by a Polypore Ca column (220 mm×4.6 mm×10 mm), both of which were
placed in an oven at 80 °C. The mobile phase used was chromatographic grade water flowing
at 0.1 mL min−1. The injection volume was 5 μL.
The sugars were identified by comparing the RT between the glucose and mannose peaks
from the patterns with the hydrolyzed xanthan gum sample. The peaks were quantified using
external aqueous solutions of glucose and mannose as standards (0.10 to 1.10 mg mL−1) to
obtain standard curves.
To determine the concentration of uronic acid, hydrolyzed polymer solutions were injected
into the HPLC adapted with an ultraviolet (UV) detector operating at a wavelength of 195 nm
(PerkinElmer 200 series) using a Polypore H pre-column (4.6 mm×30 mm×10 mm) followed
by a Polypore H column (220 mm×4.6 mm×10 mm). The columns were placed in an oven at
50 °C. The mobile phase used was aqueous H2SO4, pH 1.9 at a flow rate of 0.4 mL min−1. The
injection volume was 10 μL.
The samples were identified by comparing the retention time between the peaks of the
patterns of glucuronic acid (2.5 to 100.0 mg L−1) and pyruvic acid (2.0 to 82.0 mg L−1) with
the samples of hydrolyzed xanthan gum. The peaks were quantified using pattern curves
generated by external aqueous solution standards.

Thermal Analysis of Xanthan Gum (TG/DTG)

Non-isothermal experiments were carried out on a PerkinElmer thermogravimetric apparatus

using masses of sample of approximately 7.0 mg, a maximum temperature of 1,000 °C, and a
heating rate of 10 °C min−1 under a dynamic atmosphere of nitrogen with a flow rate of
2774 Appl Biochem Biotechnol (2014) 172:2769–2785

20 mL min−1. The thermogravimetric (TG) curves and their derivatives (DTG) were used to
determine mass loss (in percent), temperature ranges (in degree Celsius), and the maximum
temperature of thermal degradation, respectively. Both were calculated with the Pyris software
Manager.

Results and Discussion

Alternative Substrate Composition

The CGB was composed of 53.50±0.01 % volatiles, 3.40±0.01 % ash, 6.70±0.02 % total lipids,
2.71±0.03 % crude protein, and 33.69±0.02 % glycerol, including minerals, organic nitrogen,
and total lipids from biodiesel processing. The C and N sources supplied as nutrients for bacterial
growth are known to affect xanthan gum production, depending on their composition and
relative amounts. The C/N ratio in CGB is approximately 15:1. The glycerin chemical compo-
sition found here satisfies the non-limiting nitrogen concentration condition required for rapid
cell growth, while excess carbon and low nitrogen concentration are essential for the production
of xanthan gum with suitable rheological properties [26]. The nutrient source influences the
pathway by which polymers are synthesized [27]. Glycerol and lipids as free fatty acids (soaps)
are two major components in crude glycerin that contains a variety of elements such as
calcium (3–15 ppm), magnesium (1–2 ppm), phosphorous (8–13 ppm), and sulfur
(22–26 ppm), independent of the feedstock source (canola, rapeseed, and soybean) [28].
Thus, a fermentative medium richer in nutrients and micronutrients, and adaptation by the
bacterium to an alternative medium may contribute to an increase in xanthan gum production.
The composition of the CGB indicated that xanthan gum could possibly be produced from
this waste with minimal supplementation. Some authors [12, 13] produced xanthan gum from
alternative media only by supplementing the fermentation with urea (0.01 % w/v) and K2HPO4
(0.1 % w/v), resulting in lower production costs.
To produce higher amounts of xanthan gum, cell growth and polysaccharide biosynthesis
must be regulated throughout the process. The effect of fermentation time on xanthan gum
production was evaluated at different CGB concentrations (1.0, 2.0, 4.0, and 6.0 % m/v) at
1.0 vvm and 400 rpm. The experiment using 2.0 % of CGB showed the maximum accumu-
lation of xanthan gum (2.34 g L−1) after 120 h of fermentation, whereas at 1.0 % CGB, only
0.84 g L−1 of xanthan gum was accumulated. Experiments using 4 and 6.0 % of CGB resulted
in significant foaming and subsequent loss of culture medium. Thus, we used 2.0 % GRB and
a fermentation time of 120 h. These conditions are consistent with various studies on the
nutritional requirements for fermentation using Xanthomonas, aiming at the sustainability of
the process regarding cost-effectiveness of the production [3, 11, 29].

Effect of Aeration and Agitation on Xanthan Production (Y1), Biomass (Y2), and Apparent
Viscosity (Y3)

The responses to 11 tests of the crude glycerin fermentation as a function of independent

variables (X1=aeration, X2=agitation) are presented in Table 1. The coefficients of multiple
determinations (R2) give the percentage variations in the response explained by our regression
model. Thus, we can explain Y1 and Y2 variations in 96.85 % and Y3 variations in 97.06 %
(Table 2). The P value <0.05 suggests a less than 5 % chance of these variations being true.
The quadratic coefficients of X1 and X2 were significant for Y1 and Y3, while Y2 showed a
linear coefficient of X1 (P value <0.05).
Table 1 Total xanthan gum production (Y1), biomass (Y2), apparent viscosity (25 °C and 25 s−1) (Y3), glucose (Y4), mannose (Y5), glucuronic acid (Y6), pyruvic acid (Y7), average
molecular mass (Y8), and initial thermal degradation temperature (Y9) for the various fermentation conditions: aeration (X1) and agitation (X2) in a 4.5-L bioreactor at 28 °C for 120 h

Test X1 (vvm) X2 (rpm) Y1 (g L−1) Y2 (g L−1) Y3 (mPa s) Y4 (mg L−1) Y5 (mg L−1) Y6 (mg L−1) Y7 (mg L−1) Y8 (106 Da) Y9 (°C)

1 0.5 (−1) 300 (−1) 2.02 1.31 97.80 244.80 288.11 137.47 69.86 26.29 198.14
2 0.5 (−1) 700 (+1) 1.78 0.73 79.60 254.01 390.26 45.15 27.09 13.31 208.48
Appl Biochem Biotechnol (2014) 172:2769–2785

3 1.5 (+1) 300 (−1) 2.89 1.71 106.30 196.15 208.11 110.00 56.94 26.48 209.47
4 1.5 (+1) 700 (+1) 1.45 1.32 137.20 126.90 233.86 116.18 61.45 15.62 216.38
5 0.3 (−1.41) 500 (0) 4.18 1.89 165.60 0.00 750.12 71.50 149.95 17.22 194.22
6 1.7 (+1.41) 500 (0) 2.84 1.36 134.50 478.86 513.31 133.59 51.98 24.19 204.32
7 1.0 (0) 217 (−1.41) 0.87 1.23 115.20 181.60 172.60 48.03 41.20 27.79 206.52
8 1.0 (0) 783 (+1.41) 0.26 0.89 87.00 217.33 377.36 81.45 43.63 26.29 200.15
9 1.0 (0) 500 (0) 6.07 3.31 262.20 311.83 226.78 76.47 169.74 27.38 220.41
10 1.0 (0) 500 (0) 5.21 3.11 258.00 340.50 267.54 87.48 118.88 23.85 218.12
11 1.0 (0) 500 (0) 5.47 3.23 243.50 392.03 303.51 106.45 142.17 27.41 226.45
2775
2776 Appl Biochem Biotechnol (2014) 172:2769–2785

Table 2 Estimated coefficients of multiple determination (R2) and associated probability (P value) of both the
linear and quadratic effects and interaction between aeration (X1) and agitation (X2) for xanthan gum (Y1),
biomass (Y2) production, apparent viscosity (Y3), glucose (Y4), mannose (Y5), glucuronic acid (Y6), pyruvic
acid (Y7), average molecular mass (Y8), and initial thermal degradation temperature (Y9)

Term Constant X1 X1*X1 X2 X2*X2 X1*X2 R2

Y1 Coefficient 5.58 −0.34 −2.07 −0.64 −5.02 −0.59 0.968

P value 0.00 0.39 0.03 0.18 0.01 0.31
Y2 Coefficient 3.25 0.25 1.47 −0.36 −2.29 0.10 0.968
P value 0.00 0.09 0.01 0.05 0.00 0.05
Y3 Coefficient 254.58 27.47 114.68 −6.845 −163.63 24.65 0.941
P value 0.00 0.06 0.01 0.43 0.01 0.13
Y4 Coefficient 348.12 125.36 −115.68 −2.38 −155.65 −39.23 0.447
P value 0.00 0.05 0.07 0.94 0.01 0.44
Y5 Coefficient 265.95 −142.83 279.14 104.37 −77.59 −38.20 0.759
P value 0.01 0.03 0.01 0.05 0.14 0.43
Y6 Coefficient 100.13 32.84 11.69 −9.72 −26.12 −49.25 0.548
P value 0.03 0.28 0.71 0.71 0.43 0.26
Y7 Coefficient 143.59 −29.28 −51.56 −8.71 −110.11 23.64 0.789
P value 0.00 0.25 0.14 0.67 0.04 0.45
Y8 Coefficient 26.88 3.08 −7.90 −6.49 −1.57 1.06 0.649
P value 0.00 0.26 0.08 0.05 0.41 0.73
Y9 Coefficient 221.66 8.40 18.99 2.03 −14.93 −1.66 0.785
P value 0.00 0.11 0.03 0.57 0.06 0.73

The coefficient of the effect indicates how much each factor influences the responses
(Table 2). Independent variables X1 and X2 had a negative quadratic effect on Y1 and Y3,
and the transition from tested minimum to maximum values decreases these responses.
Quadratic aeration had a positive effect on Y2, whereas the linear and quadratic agitation
had negative effects.
The signs of the coefficients shown in Table 2 and the data for Y1, Y2, and Y3 in Table 1
indicate the following: When X1 is 0.5 vvm and X2 changed from 300 to 700 rpm (tests 1 to
2), the responses were reduced by 11.88, 44.27, and 18.61 %, respectively; at 1.5 vvm, this
change in agitation (tests 3 to 4) showed a reduction of 49.83 % in Y1 and 22.80 % in Y2, and
an increase of 22.52 % in Y3. When X2 changed from 0.5 to 1.5 vvm/300 rpm (tests 1 to 3),
Y1 increased by 30.10 %, Y2 by 23.39 %. and Y3 by 28.71 %; with an agitation speed of
700 rpm, this change in aeration (tests 2 to 4) decreased Y1 by 18.82 % and increased Y2 and
Y3 by 44.70 and 25.05 %, respectively. Therefore, individually, low agitation speeds contrib-
uted to increases for Y1 and Y2, while increase in aeration resulted to increases for Y2 and Y3.
The analyses in Table 1 show that the responses Y1, Y2, and Y3 are invariably reduced
when X1 and X2 are at their lowest or highest level. The test conducted at 783 rpm (test 8)
showed decreases of 84.35, 72.86, and 65.82 % in Y1, Y2, and Y3, respectively, when
compared with data from the central point (500 rpm). High agitation speeds are responsible
for hydrodynamic stress and possible damage to the cells. This death mechanism is probably
due to the shear stress caused by the blade tips of the impeller [29, 30]. A similar response was
also observed at 217 rpm (test 7), where Y1, Y2, and Y3 were reduced by approximately
83.46, 62.50, and 54.74 %, respectively, compared to the central point. The low rotor speed,
Appl Biochem Biotechnol (2014) 172:2769–2785 2777

poor mixing, and low oxygen mass transfer occurred due to the high viscosity of the broth
[31].
The efficiency of aerobic cultivation depends on cell growth. This means high consumption
rate of the carbon source in the presence of dissolved oxygen, so that there is an abundance of
electrons transported in the respiratory chain (generation of ATP). Thus, the liquid medium
must be sufficiently agitated so that oxygen is convectively transported to the cell membrane
and then penetrates the cell by simple diffusion [32]. In xanthan gum production, due to high
viscosity of the medium, it is necessary to transmit high frequency agitation so that air bubbles
get smaller diameters and do not coalesce.
It is interesting to notice that the biomass profile is similar to the one of xanthan production,
showing that polymer production is related to the biomass concentration and that both of them
are functions of the agitation speed and aeration (Table 1). Therefore, there is a good chance
that this process with a high biomass concentration may ensure a high xanthan production. It
can also be seen that there is an increasing death rate as revolution speed increases.
The aeration and agitation conditions become important to prevent anaerobic conditions
due to the high viscosity of the medium that results from xanthan gum formation.
Previous studies [8] have evaluated the production of xanthan gum from glucose by
X. campestris ATCC 1395 at 100 to 800 rpm and 1 vvm. Production of xanthan gum nearly
doubled when the agitation speed was increased from 100 (3.0 g L−1) to 600 rpm (6.3 g L−1)
and the biomass concentration was maximized (2.8 g L−1) at 600 rpm. The xanthan gum
production from date waste by X. campestris PTCC 1473 evaluated at 200, 300, and 400 rpm
in 1.0 vvm [33] was maximized at 400 rpm (6.93 g L−1) and was minimized at 200 rpm
(1.55 g L−1).
A reduction in aeration from 1.7 to 1.0 vvm/500 rpm (tests 6 to 9, Table 1) increased Y1,
Y2, and Y3 by 53.21, 58.92, and 48.70 %, respectively. However, reducing aeration from 1.0
to 0.3 vvm (tests 9 to 5) reduced Y1 by 31.13 %, Y2 by 42.90 %, and Y3 by 36.84 %.
The average amount of xanthan gum from CGB (5.58 g L−1) was 35.4 and 30.0 times
higher than the production in the shaker (180 rpm/28 °C) using glycerol and two strains of
Xanthomonas sp. (C1 and C9) in a previous study [20]. Likewise, the average apparent
viscosity of xanthan gum derived from CGB (254.56 mPa s) was 1.82 times higher than the
viscosity of xanthan gum produced by the C1 strain.
After disposing all the factors and interactions of factors that were not significant (P value
>0.05) due to lack of fit, an analysis of variance was conducted and coefficients (R2) of the
statistical models were determined. The analysis of variance (Table 3) shows that the quadratic
models adjusted for Y1, Y2, and Y3 (Eqs. 3, 5, and 7) were significant with 95 % confidence

Table 3 R2 of polynomial models, P value, and F value obtained from ANOVA for xanthan gum production
(Y1), biomass (Y2), apparent viscosity (Y3), glucose (Y4), mannose (Y5), glucuronic acid (Y6), pyruvic acid
(Y7), average molecular mass (Y8), and maximum thermal degradation temperature (Y9)

Y1 Y2 Y3 Y4 Y5 Y6 Y7 Y8 Y9

Regression R2 0.97 0.97 0.94 0.44 0.76 0.55 0.79 0.65 0.78
F 20.48 18.61 22.01 0.51 2.10 0.81 2.51 1.22 2.44
F(5,5)a 5.05
Lack of fit P 0.44 0.14 0.19 0.06 0.06 0.53 0.34 0.12 0.05
F 1.39 6.09 4.26 18.74 14.59 1.03 2.08 3.83 3.25
F(3,2)a 19.16
a
Values obtained from the F distribution table at 5 %
2778 Appl Biochem Biotechnol (2014) 172:2769–2785

and predictive, once F-calculated regression was approximately 4.06, 4.08, and 4.36 times
higher than the F-tabulated regression for Y1, Y2, and Y3, respectively. In addition, the value
of the determination coefficient (R2), which assesses the fit of the model, was 0.968 for Y1,
0.968 for Y2, and 0.971 for Y3. These values indicate that the models for Y1, Y2, and Y3 did
not explain only 3.15, 3.14, and 2.94 % of the total variations, respectively.
The final equations for Y1, Y2, and Y3 were derived by applying the response surface
methodology. These equations, along with the response surface set in canonical form (Y b ) are
given below:
The production of xanthan gum (Y1):

Y1 ¼ 5:583 − 0:169X1 − 1:036X12 − 0:318X2 − 2:509X22 − 2:299X1X2 ð3Þ

b ¼ 5:589 − 0:642W12 − 4:187W22

Y1 ð4Þ
The biomass (Y2):

Y2 ¼ 3:247 þ 0:124X1 − 0:736X12 − 0:183X2 − 1:146X22 − 0:05X1X2 ð5Þ

b ¼ 3:259 þ 0:737W12 − 1:147W22

Y2 ð6Þ
The apparent viscosity (Y3):

Y3 ¼ 254:58 þ 13:735X1 − 57:341X12 − 3:423X2 − 81:815X22 þ 12:325X1X2 ð7Þ

b ¼ 255:403 − 52:209W12 − 86:951W22

Y3 ð8Þ
The negative signals to the roots λ1 (−0.641, −0.737, and −52.209) and λ2 (−4.187,
−1.147, and −86.951) showed that the responses Y1, Y2, and Y3 have a maximum point
confirmed by the surface shown in Fig. 1.
Xanthan production was maximized (5.59 g L−1) at an aeration of 0.97 vvm and an
agitation speed of 497.76 rpm, biomass (3.26 g L−1) at 1.05 vvm and 484.75 rpm, and
apparent viscosity (255.40 mPa s) at 1.05 vvm and 499.40 rpm. These results agree with
those shown in Table 1, which show that the maximum production was obtained at the central
point (1.0 vvm and 500 rpm).
Figure 2 shows that the apparent viscosity of all obtained biopolymers decreases as the
shear rate increases. According to a previous study [34], this behavior is characteristic of
pseudoplastic fluids and typical of xanthan gum.

Effect of Aeration (X1) and Agitation (X2) on the Chemical Composition of Xanthan Gum
(Y4 to Y7)

The chromatograms of the glucose, mannose, glucuronic acid, and pyruvic acid patterns are
shown in Figs. 3a and 4a. The xanthan compositions were quantified by calibration curve
(peak area×standards concentration in milligrams per milliliter), which yielded the following
equations: Glucose=(peak area+13,277)/305,515(R2 =0.987), Mannose=(peak area−5,628)/
39, 606(R 2 = 0.998), Glucuronic acid = (peak area − 58.65)/1, 356(R 2 = 0.998), and
Pyruvic acid=(peak area−2,382)/17,615(R2 =0.999).
Appl Biochem Biotechnol (2014) 172:2769–2785 2779

Fig. 1 Response surface plots of coded variables versus aeration (X1) and agitation (X2) for xanthan production
(a), biomass concentration (b), and apparent viscosity of xanthan solution (c)

The chemical composition was characterized by four responses: glucose (Y4), mannose
(Y5), glucuronic acid (Y6), and pyruvic acid (Y7), shown in Table 1 and Figs. 3 and 4. The
estimated effects are reported in Table 2 and indicate that aeration linearly influences the
glucose and mannose concentrations (P value <0.05) and quadratically influences the mannose
concentration (P value <0.05). Agitation quadratically influences the glucose and pyruvic acid
concentrations (P value <0.05), while it linearly influences the mannose concentration.

Fig. 2 Apparent viscosity of xanthan gum solutions (0.5 %) derived from sucrose (control) and CGB at 25 °C as
a function of shear rate
2780 Appl Biochem Biotechnol (2014) 172:2769–2785

Fig. 3 Sugar standards HPLC-RI chromatograms (a) and sugar composition of xanthan gums derived from
GRB by Xanthomonas campestris mangiferaeindicae 2103 at the central point (500 rpm and 1.0 vvm, test 9) (b),
at the highest levels of aeration and agitation (700 rpm and 1.5 vvm), (c) and at minimum levels (0.5 vvm and
300 rpm) (d)

The signs of the coefficients isolated for X1 and X2 for Y4 (Table 2) indicate that the
increase in aeration or agitation results in the increase of glucose concentration of the polymer.
This behavior was observed when the agitation speed was maintained at 500 rpm and the
aeration was increased from 0.3 vvm (test 5) to 1.5 vvm (test 6). While these conditions
increased the glucose concentration by 100 % (Table 1), the aeration rate increase under
constant agitation reduced the glucose concentration in other instances. Therefore, aeration and
agitation have a negative effect on the glucose concentration.
The signs of the coefficients isolated for X1 and X2 for response Y5 (Table 2) indicate that
decreasing the aeration rate or increasing the agitation speed increases the mannose concen-
tration in the polymer. Y5 increases by 26.17 % when X1 is 0.5 vvm and X2 changed from
300 to 700 rpm (tests 1 to 2), by 11.01 % when X1 is 1.5 vvm (tests 3 to 4), by 27.80 % when
X2 is 300 rpm and X1 changed from 0.5 to 1.5 vvm (tests 1 to 3), and by 40.07 % when X2 is
700 rpm (tests 2 to 4).
The coefficients calculated for Y6 (Table 2) reveal that the variables and their interactions
did not affect the response of Y6 at the 95 % confidence level.
The coefficients for Y7 (Table 2) showed that the response was reduced by 61.22 % when
X1 is 0.5 vvm and X2 changed from 300 to 700 rpm (tests 1 to 2) and by 7.48 % when X1 is
1.5 vvm (tests 3 to 4). Therefore, increase in agitation results in increase of the polymer
pyruvic acid.
The monosaccharide composition of xanthan gum derived from CGB near the optimized
point (500 rpm and 1.0 vvm) is shown in Figs. 3b and 4b and yielded a molar ratio of
Appl Biochem Biotechnol (2014) 172:2769–2785 2781

Fig. 4 HPLC-UV chromatograms of acid standards (a) and acids of xanthan gums obtained from GRB by
Xanthomonas campestris mangiferaeindicae 2103 at the central point (500 rpm and 1.0 vvm, test 9) (b), at the
highest levels of aeration and agitation (700 rpm and 1.5 vvm, test 4) (c), and at minimum levels (0.5 vvm and
300 rpm, test 1) (d)

3.3:2.8:1.0:1.6 (glucose/mannose/glucuronic acid/pyruvic acid) (Table 1). Thus, the primary

structure likely consists of repeated heptasaccharide units, comprising three glucose, three
mannose, and one glucuronic acid units, in addition to pyruvic acid substituents. These
conditions differ from those obtained under extreme aeration and agitation conditions (test 1
and test 4) shown in Figs. 3c, d and 4c, d. The glucose/mannose/glucuronic acid/pyruvic acid
ratios under extreme aeration and agitation were 1.8:2.1:1.0:0.5 and 1.1:2.0:1.0:0.5, respec-
tively. Previous authors demonstrated [2, 3] that the xanthan primary structure is a pentasac-
charide composed of two glucose units, two mannose units, and one glucuronic acid unit.
Variations in the microbial polymer composition occur frequently depending on the strain and
medium employed. Typically, the xanthan gum produced from sucrose by different species of
Xanthomonas contains glucose, mannose, and glucuronic acid [3].
The concentration of pyruvate substituent may vary depending on the conditions of the
fermentation process because the rheological properties of xanthan gum are directly related to
the amount of these substituents [35]. The xanthan gum derived from CGB by X. campestris
mangiferaeindicae 2103 showed concentration of pyruvate ranging from 3.78 to 20.16 %
(Table 1 and Fig. 4) depending on the aeration and agitation conditions. Most xanthan gums
produced by X. campestris show higher pyruvate concentrations than those reported, ranging
between 4.4 and 7.6 % [36, 37]. Xanthan gum samples with pyruvate concentrations greater
than 4 % are more viscous than those with lower concentrations (0.5 to 3 %) [37]. Therefore,
the pyruvic acid concentration directly affects the apparent viscosity of the xanthan gum
solutions derived from CGB (R2 =0.80). This effect can be attributed to the carbon source and/
2782 Appl Biochem Biotechnol (2014) 172:2769–2785

or formation of macromolecules by pyruvate, which promotes the affinity between the

polymer chains according to a previous study [38].
The results of the ANOVA are reported in Table 3 and indicate that adjustments of the
possible polynomial models that describe the glucose, mannose, and glucuronic and pyruvic
acid concentrations were not valid (R2 <0.8), indicating that the models did not explain
approximately 20 % of the total variations. Furthermore, the models are not predictive because
the F-calculated regression was approximately 9.90, 2.40, 6.23, and 2.01 times lower than the
F-tabulated regression for Y4, Y5, Y6, and Y7, respectively.

Effect of Aeration (X1) and Agitation (X2) on the Average Molecular Mass (Y8) and Thermal
Stability of Xanthan Gum (Y9)

The average molecular mass of xanthan gum derived from CGB using strain 2103 (Table 1 and
Fig. 5) was determined and compared with the xanthan gum Sigma (reference). The TG/DTG
curves (Fig. 6) illustrate the thermal decomposition mechanism of xanthan gum derived from
CGB produced at the central point, and extreme aeration and agitation conditions (tests 1 and
4) compared with xanthan gum Sigma (reference).
The influence of the independent variables X1 and X2 on Y8 and Y9 revealed a negative
effect of agitation on Y8 and a positive effect of aeration on Y9 (P value <0.05).
Analysis of the data for Y8 (Table 1) shows reductions of 49.37 and 41.01 % when
changing the agitation speed from 300 to 700 rpm at 0.5 vvm (tests 1 to 2) and 1.5 vvm (tests

Fig. 5 GPC HPLC-RI chromatograms showing the distribution molecular mass of xanthan gum derived from
CGB by Xanthomonas campestris mangiferaeindicae 2103 at the central point (500 rpm and 1.0 vvm, test 9) (a),
at the highest levels of agitation and aeration (700 rpm and 1.5 vvm, test 4) (b), and at the minimum levels
(300 rpm and 0.5 vvm, test 1) (c), and xanthan gum Sigma (reference) (d)
Appl Biochem Biotechnol (2014) 172:2769–2785 2783

Fig. 6 TG (solid line) and DTG (broken line) curves of xanthan gum obtained from CGB by Xanthomonas
campestris mangiferaeindicae 2103 at the central point (500 rpm and 1.0 vvm, test 9) (a), under higher levels of
agitation and aeration (700 rpm and 1.5 vvm, test 4) (b), and under minimum levels (300 rpm and 0.5 vvm, test 1)
(c), and xanthan gum Sigma (reference) (d) at 10 °C min−1

3 to 4), respectively. However, increasing the aeration from 0.5 to 1.5 vvm increased Y9 by
approximately 5.41 and 3.65 % at 300 rpm (tests 1 to 3) and 700 rpm (tests 2 to 4),
respectively. Therefore, low agitation speeds (X2) may increase the molecular mass of polymer
(Y8), while high aeration increased the resistance to thermal degradation of the xanthan gum
(Y9).
The composition of the fermentation medium, the Xanthomonas strain, and the operating
conditions significantly impact the molecular structure of xanthan gum, producing polysac-
charides with molecular masses that can vary from 2.0 to 20×106 Da [3]. Thus, the different
average molecular masses shown in Table 1 and Fig. 5 can be attributed to the substrates used
and the selected operating conditions. Therefore, the maximum molecular mass of xanthan
gum derived from CGB (27.38×106 Da) is lower than that of the xanthan gum obtained from
Sigma (49.44×106 Da), which was produced on an industrial scale using glucose as the
substrate.
Table 1 also shows that the higher viscosity values are directly associated (R2 =0.88) with
higher molecular masses for xanthan gum with pyruvic acid concentrations greater than 10 %.
Xanthan gum with high molecular mass and tertiary structure affects the viscosity of the
biopolymer aqueous solution [27].
The TG/DTG curves (Fig. 6) illustrate the thermal decomposition mechanism of xanthan
gum derived from CGB at the central point and extreme aeration and agitation conditions (tests
1 and 4). These values were compared with the xanthan gum obtained from Sigma (reference).
An analysis of Fig. 6 indicates that the TG/DTG curves of xanthan gum derived from CGB
showed a behavior similar to that of the xanthan gum Sigma, where two thermal events were
observed. The first mass loss (24.8–68.7 °C, 23.6–68.7 °C, 22.5–82.6 °C, and 21.8–78.9 °C)
corresponds to 15.71, 15.16, 12.05, and 12.90 % of the xanthan gum produced in tests 9, 1,
2784 Appl Biochem Biotechnol (2014) 172:2769–2785

and 4 and the xanthan gum Sigma, respectively. This event is related to the dehydration of the
polymer. The presence of water in the lyophilized samples may be attributed to rapid water
uptake during weighing due to the increased hygroscopic character of the lyophilized poly-
saccharide. According to a previously published study [39], xanthan gum absorbs water due to
the presence of polar groups in its structure, especially the –OH groups. The second mass loss
corresponds to the degradation of the polymer chain [40].
The xanthan gum derived from CGB produced in test 9 degraded between 220.41 and
279.56 °C, and the xanthan gum Sigma degraded between 200.71 and 299.60 °C with mass
losses of 49.92 and 47.30 %, respectively. The xanthan gum produced in test 1 (0.5 vvm and
300 rpm) degraded between 198.14 and 292.30 °C, and that produced in test 4 (1.5 vvm and
700 rpm) degraded between 216.38 and 274.60 °C with mass losses of 48.63 and 52.61 %,
respectively.
The xanthan gum produced at the central point (220.41 °C) was more thermally stable than
the xanthan gum obtained from test 1 (198.14 °C), test 4 (216.38 °C), and the xanthan Sigma
(200.71 °C), as indicated by the initial thermal degradation temperatures. Therefore, xanthan
produced at the point of optimization (1.0 vvm and 500 rpm) can be used in processes that
require high temperatures.
The results of the ANOVA are reported in Table 3 and indicate that adjustments of the
possible polynomial models that describe the responses Y8 and Y9 were not valid (R2 <0.8),
indicating that the models did not explain approximately 20 % of the total variations.
Furthermore, the models are not predictive because the F-calculated regression was 4.13 and
2.07 times lower than the F-tabulated regression for Y8 and Y9, respectively.

Conclusions

Low agitation speeds increase the production of xanthan gum, biomass concentration, apparent
viscosity of xanthan solutions, glucose and pyruvic acid concentrations in the xanthan chain,
and molecular mass. On the other hand, increasing the agitation speed contributes to the
increase in mannose concentration in the chain.
Increasing the aeration rate contributes to an increase in the biomass concentration, apparent
viscosity of the xanthan solutions, and the resistance to thermal degradation. However,
decreasing the aeration contributes to increased xanthan production and higher concentrations
of glucose and mannose in the polymeric chain.
Optimization procedures of CGB fermentation in a bioreactor at 28 °C for 120 h indicated
that 0.97 vvm at 497.76 rpm resulted in a xanthan gum production of 5.59 g L−1 with
255.40 mPa s at 0.5 % (m/v), 25 °C, and 25 s−1.
Overall, our findings indicate that the studied date crude glycerin co-products have strong
potential and promising properties that can open new pathways for the efficient and cost-
effective production of xanthan gum, which is an alternative to the disposal of the residue. As
such, they can be considered strong candidates for future industrial and commercial applica-
tions related to xanthan gum.

References

1. Palaniraj, A., & Jayaraman, V. (2011). Journal of Food Engineering, 106, 1–11.
2. Rosalam, S., & England, R. (2006). Enzyme and Microbial Technology, 39, 197–207.
3. Garcia-Ochoa, F., Santos, V. E., Casas, J. A., et al. (2000). Biotechnology Advances, 18, 549–579.
Appl Biochem Biotechnol (2014) 172:2769–2785 2785

4. Ben Salah, R., Chaari, K., Besbe, S., et al. (2010). Food Chemistry, 121, 627–633.
5. Jasson, P. E., Kenne, L., & Lindeberg, B. (1975). Carbohydrate Research, 45, 275–282.
6. Duta, F. P., França, F. P., & Lopes, L. M. A. (2006). Electronic Journal of Biotechnology, 9, 393–399.
7. Nicolaus, B., Kambourova, M., & Oner, E. T. (2010). Environmental Technology, 31, 1145–1158.
8. Papagianni, M., Psomas, S. K., Batsilas, L., et al. (2001). Process Biochemistry, 37(73), 80.
9. Kerdsup, P., Tantratian, S., Sanguandeekul, R., et al. (2011). Food Bioprocess Technology, 4, 1459–1462.
10. Bilanovic, D., Chang, F., Isobaev, P., et al. (2013). Biomass and Bioenergy, 35, 2683–2689.
11. Druzian, J. I., & Pagliarini, A. P. (2007). Ciência e Tecnologia de Alimentos, 27, 26–31.
12. Nery, T. B. R., Brandão, L. V., Esperidião, M. C. A., et al. (2008). Quimica Nova, 39, 1937–1941.
13. Brandão, L. V., Esperidião, M. C. A., & Druzian, J. I. (2010). Polímeros, 20, 1–6.
14. Ayoub, M., & Abdullah, A. Z. (2012). Renewable and Sustainable Energy Reviews, 16, 2671–2686.
15. Anand, P., & Saxena, R. K. (2011). New Biotechnology, 29, 199–205.
16. Çelik, E., Ozbay, N., Oktar, N., et al. (2008). Industrial and Engineering Chemistry Research, 4, 2985–2990.
17. Zhu, Y., Li, J., Tan, M., et al. (2010). Bioresource Technology, 101, 8902–8906.
18. Teodoro, J. C., Baptista-Neto, A., Araujo, M. L. G. C., et al. (2010). Brazilian Journal of Chemical
Engineering, 27, 499–506.
19. Rywińska, A., & Rymowicz, W. (2011). Chemistry Papers, 65, 119–123.
20. Reis, E. C., Alemida, M., Cardoso, J. C., et al. (2010). Macromolecular Symposia, 296, 347–353.
21. Zhao, L., He, Y., Deng, X., et al. (2012). Molecules, 17, 3618–3629.
22. Galindo, E., Salcedo, G., & Ramirez, M. E. (1994). Applied Microbiology and Biotechnology, 40, 634–637.
23. Horwitz, W. (1997). Official methods of analysis of AOAC International, 16th ed., Gaithersburg, Maryland:
Association of Official Analytical Chemists International.
24. Bligh, E. G., & Dyer, W. J. (1959). Canadian Journal of Biochemistry and Physiology, 37, 911–917.
25. Myers, R. H., & Montgomery, D. C. (2002). Response surface methodology, process and product optimi-
zation using designed experiments (2nd ed.). New York: Wiley.
26. Sutherland, I. W. (1996). Biotechnology, 6, 613–657.
27. Uma Shankar, M. K. S., Uma Shaanker, R., Ganeshaiah, K. N., et al. (1996). Economic Botany, 50, 270–279.
28. Thompson, J. C., & He, B. B. (2006). Applied Engineering in Agriculture, 22, 261–265.
29. Casas, J. A., Santos, V. E., & García-Ochoa, F. (2000). Enzyme and Microbial Technology, 26, 282–291.
30. Silva-Santisteban, B. O. Y., & Filho, F. M. (2004). Enzyme and Microbial Technology, 36, 717–724.
31. Peters, H. U., Herbst, H., Hesselink, P. G. M., et al. (1989). Biotechnology and Bioengineering, 34, 1393–
1397.
32. Cherry, R. S., & Papoutsakis, H. (1986). Bioprocess Engineering, 1, 29–41.
33. Moshaf, S., Hamidi-Esfahani, Z., & Azizi, M. H. (2011). World Academy of Science Engineering and
Technology, 57, 521–524.
34. Kalogiannis, S., Iakovidou, G., Liakopoulou-Kyriakides, M., et al. (2003). Process Biochemistry, 39, 249–
256.
35. Flores-Candia, J. L., & Decker, W. D. (1999). Biotechnology Progress, 15, 531–538.
36. Orentas, D. G., Sloneker, J. H., & Jeanes, A. (1963). Canadian Journal of Microbiology, 9, 427–430.
37. Shatwell, K. P., Sutherland, I. W., & Ross-Murphy, S. B. (1990). International Journal of Biological
Macromolecules, 12, 71–78.
38. Smith, I. H., Symes, K. C., Lawson, C. J., et al. (1981). International Journal of Biological Macromolecules,
3, 129–134.
39. Faria, S., Petkowicz, C. L. O., Moraes, S. L., et al. (2011). Carbohydrate Polymer, 86, 469–476.
40. Zohuriaan, M. J., & Shokrolahi, F. (2004). Polymer Testing, 23, 575–579.
Copyright of Applied Biochemistry & Biotechnology is the property of Springer Science &
Business Media B.V. and its content may not be copied or emailed to multiple sites or posted
to a listserv without the copyright holder's express written permission. However, users may
print, download, or email articles for individual use.