Contents lists available at ScienceDirect

Biochemical Engineering Journal
journal homepage: www.elsevier.com/locate/bej
Regular article

Rheological characteristics of filamentous cultivation broths and suitable
model fluids
Bliatsiou C.
a ,
*
, 1
, Schrinner K.
b , c , 1
, Waldherr P.
a , 1
, Tesche S.
b , c
, Böhm L.
a
, Kraume M.
a
, Krull R.
b , c
a
Chair of Chemical and Process Engineering, Technische Universität Berlin, Berlin, Germany
b
Institute of Biochemical Engineering, Technische Universität Braunschweig, Braunschweig, Germany
c
Center of Pharmaceutical Engineering (PVZ), Technische Universität Braunschweig, Braunschweig, Germany
HIGHLIGHTS
• Broths of A.niger , L.aerocolonigenes , A.namibiensis exhibit non-Newtonian rheology.
• Suitable rheometer geometry and measuring method varies for each cultivation broth.
• Broth viscosity and flow consistency factor increase with biomass growth.
• Broth viscosity increases with increased pellet roughness and decreased compactness.
• Xanthan gum solutions can be used as model fluids for A.niger and A.namibiensis .
ARTICLE INFO
Keywords:
Rheology
Filamentous cultivation broth
Cell morphology
Model fluid system
Non-newtonian behavior
ABSTRACT
The rheological characteristics of cultivations with filamentous microorganisms define many other crucial
process parameters in the system, such as the cell morphology and the productivity. The present study aims to
determine the rheological changes of broths during batch cultivations of three filamentous microorganisms: the
fungus Aspergillus niger , and the two actinomycetes Lentzea aerocolonigenes and Actinomadura namibiensis . Focus
is given on establishing measurement methodologies to characterize these heterogeneous filamentous systems.
For the rheological characterisation the Ostwald-de Waele power law approach is used. All examined biological
filamentous broths exhibit a non-Newtonian shear thinning behavior. The effect of biomass concentration and
cell morphology on the rheology of the broth is further enlightened. Finally, the rheology of the cultivation
broths is compared with that of various model fluids to test the suitability of the last ones to mimic rheologically
the biological broths in further technical applications.
1. Introduction
Filamentous microorganisms are commercially used as cell factories
in multiple sectors such as pharmaceutical, food and beverage, paper
and pulp or textile industry [ 1 , 2 ]. Knowledge of the broth properties,
specifically the broth rheology is of paramount importance for the ef-
ficient performance of the cultivation process [ 3 ]. The rheological
properties influence turbulence characteristics of the fluid and, there-
fore, heat and mass transfer processes [ 4 ], the mixing behavior [ 5 ] and
the fluid mechanical stress [ 6 ], defining the productivity of the culti-
vation and thus, the overall performance of the cultivation process.
Finally, scale-up tasks and downstream processes also require a good
understanding of rheological properties [ 7 ].
During cultivation, rheology can be drastically affected by the
consumption of viscous substrates, the growth of filamentous micro-
organisms or the production of viscous products (e.g., biopolymers).
This work deals exclusively with the rheological changes linked to
biomass growth and the morphology of filamentous microorganisms. It
is generally acknowledged that the non-Newtonian flow behavior of
filamentous cultivation broths is caused by the presence of micro-
organisms with complex cell morphology [ 8 ]. The formation of highly
branched mycelia can increase the viscosity and often leads to non-
https ://doi.org/10.1016/j.bej.2020.107746
Received 11 May 2020; Received in revised form 25 July 2020; Accepted 28 July 2020
Abbreviations: CMC, carboxymethyl cellulose solution; DSMZ, German collection of microorganisms and cell cultures; PEG, polyethylene glycol; PAA, poly-
oxpolyacrylamide

⁎
Corresponding author at: Technische Universität Berlin, Fachgebiet Verfahrenstechnik, ACK 7, Ackerstraße 76, 13355, Berlin, Germany.
E-mail address: [email protected] (C. Bliatsiou).
1
C. Bliatsiou, K. Schrinner and P. Waldherr should be considered joint first authors.
Biochemical Engineering Journal 163 (2020) 107746
Available online 03 August 2020
1369-703X/ © 2020 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY license
(http://creativecommons.org/licenses/by/4.0/).
T

Newtonian behavior of aqueous nutrient solutions [ 9 ].
Shear rate dependent viscosity is usually handled using the apparent
viscosity
= µ app

(1)
μ

app
can be determined by measuring shear stress
as a function of
shear rate . In stirred tank reactors and impeller rheometers, a broad
distribution of shear rates can be applied.
For the laminar flow regime, Metzner and Otto [ 10 ] developed a
simple approach to calculate a representative shear rate by using a
constant K

MO
depending on the geometry of the stirring set-up and the
stirring rate N :
= K N
MO

(2)
The non-Newtonian behavior of filamentous cultivation broths is
most commonly described using the Ostwald-de Waele power law ap-
proach [ 11–14 ]

= K n

(3)
due to its simplicity and since it has proven to adequately describe flow
curves over a wide range of shear rates [ 7 ]. Furthermore, it is the most
frequently and successfully used model in transport correlations [ 7 ].
The flow consistency factor K stands for the inner friction and the flow
behavior index n can be interpreted as a qualitative information on the
stability of the fluid’s inner structure. Other rheological models used in
literature to describe filamentous cultivation broths include the Her-
schel-Bulkley correlation with the yield stress τ

0
, the Bingham (ideal
plastic, n = 1) and Casson model (shear thinning, 1/3 < n < 2/3). The
choice of the rheological model is mostly subjective and depends on the
experimental condition. Rheological models must be understood as
empirical correlations and not as physical laws [ 15 ].
Rheology of the broth is influenced decisively by the biomass con-
centration and the morphological state of the mycelium [ 5 ]. Depending
on the cultivation conditions (e.g., pH-value, temperature, osmolality,
reactor geometry, etc.), filamentous microorganisms grow in the fol-
lowing main morphological forms: (a) homogeneously dispersed my-
celia with each individual hypha being surrounded by the cultivation
medium, (b) pellet-like bio-agglomerates, in which the hyphal ag-
glomerate assembles mostly into a spherical form, and (c) all inter-
mediate types of clumps. The morphological state of the microorgan-
isms affects significantly the productivity of the cultivation. Whether
pellet growth or cultivation of freely dispersed mycelia is preferred,
depends on the microorganism and strain, the desired product and the
specific application [ 16 ]. Extensive research was carried out to under-
stand, control and optimize cell morphology of filamentous micro-
organisms to increase the product concentrations [ 17 ].
Several authors studied the influence of both the biomass
concentration and the mycelia morphology on the Ostwald-de Waele
parameters for different cultivation systems. In those works, the bio-
mass was usually expressed as Biomass Dry Weight concentration ( BDW )
and the morphology was expressed in Euclidian or fractal terms of
clump roughness and compactness [ 18 ], diameter of mycelial aggregate
[ 19 ], projected area [ 20 ], equivalent diameter [ 12 ] or the combination
of these shape descriptors in the non-dimensional Morphology number,
fractal quotient or lacunarity [ 15 ]. These parameters are used for the
description of different aspects of the pellet morphology. The roughness
describes the irregularity of an object perimeter, which is obtained from
circularity measurements, whereas the compactness is a measure for the
voidage in a pellet or mycelial aggregate (for details see [ 17 ] and [ 21 ]).
An equivalent diameter usually describes the diameter of a circle with
the same area (or volume) as the pellet [ 22 ]. The equivalent diameter
together with the projected area (area of a two-dimensional re-
presentation of the pellet) are both measures for the size of pellets [ 19 ].
The studies mentioned above focused on the dependency between the
rheological data and BDW as well as the chosen morphological para-
meter. Results strongly vary according to the investigated micro-
organism and cultivation conditions.
Rheological properties are intrinsic parameters of the fluid and
should not be influenced by the measurement system. Nevertheless, due
to the highly heterogeneous nature of cultivation broths, conducting
rheological measurements is challenging and no standard methodology
exists [ 7 ]. Measurements are mostly done offline, although approaches
for online measurements have been proposed [ 23–26 ]. In the literature,
several measuring geometries were used. Conventional rotational
viscometer set-ups such as double gap cells [ 7 , 24 , 27 , 28 ], concentric
cylinders [ 7 , 24 , 29 ] or cone and plate [ 12 ]. However, there are several
typical problems associated with these rheometers. The biopellet
structure may be damaged or altered, as the size of the agglomerates
often is of the same size or larger than the annulus/gap size of the
instrument. Also the heterogeneous nature of the suspension can lead to
erroneous measurements due to sedimentation or aligning of pellets
[ 5 ]. Bongenaar et al. [ 30 ] were the first ones employing the impeller
viscometer to cultivation broths, which since then has been employed
by several authors [ 7 , 15 , 24 , 31 ]. Using this indirect impeller method,
the investigated range of shear rate has to be chosen carefully, as the
governing equations are theoretically only valid in the laminar flow
regime. However, also a lower limit of shear rates exists to ensure a
homogeneous suspension without pellet sedimentation. In an extensive
study for the comparison among various rheometer geometries, it was
suggested that eventually different rheometers should be used to
characterize successfully a cultivation broth at the different stages of
the cultivation process [ 32 ].
The rheology of a cultivation broth defines the fluid dynamic con-
ditions in the bioreactor. Studying the hydrodynamic behavior and the
mass transfer of various bioreactor configurations can become
Nomenclature
Latin Symbols
d
av
[m] Average Diameter based on the Maximum Feret Diameter
BDW [g/L] Biomass Dry Weight concentration
c [g/L] or [g/kg] or [% w/w] Concentration
t [s] Cultivation Time
= El [ ]
N d µ / ( )
N 1 ( ) ( )
2

Elasticity Number
where N

1
first normal stress difference,
fluid density
d impeller diameter

n

[-] Flow Behavior Index
K [Pa∙s

n
] Flow Consistency Factor
s [m] Gap Size
N [s
−1
] Impeller Speed
K MO

[-] Metzner-Otto Constant
C

K
or C
n
[-] Parameter
Re [-] Reynolds Number
Greek Symbols
µ app

[Pa∙s] Apparent Viscosity

[ ]

Cone Angle

[ ]
eff

Effective Volume Fraction

[µS / cm]

Electrical Conductivity

µ [Pa s]

Newtonian Dynamic Viscosity

[ ]

Parameter, exponent in the power law approach

[s ]
1

Shear Rate

[Pa]
0

Yield Stress

C. Bliatsiou, et al. Biochemical Engineering Journal 163 (2020) 107746
2

Table 1
Employed biological and model fluid media.
System Description
Cultivation broth Cultivation Biochemical variable Cell morphology Cultivation type/ Period [h]
Aspergillus niger AS0 400 mosmol/kg (without NaCl addition), inoculum concentration 10
6
spores/mL Exclusively in pellet form with dense core and a thin fluffy region
surrounding the core, big pellets
Stirred tank bioreactor/
53.5 h
AS1 400 mosmol/kg (without NaCl addition), inoculum concentration 3.9 × 10
5
spores/mL Exclusively in pellet form with dense, compact core without fluffy region
surrounding the core, big pellets
Stirred tank bioreactor/
68.5 h
AS2 1500 mosmol/kg ( c
NaCl
= 32.2 g/L), inoculum concentration 10
6
spores/mL Mainly in mycelial form, presence of pellets of losing structure, big bio-
agglomerates
Stirred tank bioreactor/
106.5 h
AS3 2500 mosmol/kg ( c
NaCl
= 64.4 g/L), inoculum concentration 10
6
spores/mL Mainly in small non-spherical, elongated clumps with high compactness Stirred tank bioreactor/
93.5 h
AS4 pH-shift from 3 to 5 from cultivation time 24–29 h, 400 mosmol/kg (without NaCl
addition), inoculum concentration 10
6
spores/mL
Mainly in small mycelial forms, presence of non-spherical, elongated
compact clumps
Stirred tank bioreactor/47 h
Lentzea aerocolonigenes LA – Hairy pellets of medium size and many small mycelia Shaking flask/240 h
Actinomadura namibiensis AC – Mixture of pellets, clumps and dispersed mycelia, pellets with irregular
shape
Shaking flask/240 h
Model fluid medium Concentration, c [g/kg] Newtonian dynamic viscosity, μ [Pa∙s] Rheological behavior
Inverted sugar syrup solution 10 - 1000 9.7
10

−4
- 0.50 Newtonian
Glycerin solution 10 - 1000 9.4
10

−4
- 1.313
Luviskol ® K90 solution 10 - 200 2.64
10

−3
- 1.704
Polyethylene oxide-PEG solultion 20−500 – shear thinning
Xanthan gum solution 0.5−10 –
Carboxymethyl cellulose-CMC solution 0.1−10 –
C. Bliatsiou, et al. Biochemical Engineering Journal 163 (2020) 107746
3

challenging, since the real biological process material is often difficult
to work with. Biological broths are usually opaque, prohibiting optical
accessibility; they are stable only under specific temperature and
pressure conditions, and allowed to be used under specific safety
standards, which can consequently increase the cost of the investiga-
tions [ 33 ]. Therefore, such studies are usually facilitated by the use of
non-biological based model systems, i.e., model particulate systems and
model fluids. The use of suitable model systems allows faster, re-
producible and thus, more cost-effective experimentations towards the
optimization of technical relevant reactors. Experiments with the actual
biological material are normally carried out only selectively to confirm
the laboratory results obtained with model fluids.
A wide range of model fluids is reported in the literature. The model
fluids used in laboratory must behave rheologically in a manner re-
presentative of the actual process fluid. They often need to be trans-
parent for optical accessibility. Low cost and easy to follow safety
standards, as well as a safe and economical disposal are also required.
Characteristics such as the stability of the fluid, slow decomposition and
temperature sensitivity need to be considered [ 33 ]. Model fluids have
been used in investigations of flow visualization, power input, fluid
mechanical stress, mixing times, solid-liquid mixing (solid suspension
and distribution), liquid-liquid dispersion and gas-liquid mixing (gas
hold-up, volumetric mass transfer coefficient, bubble size and specific
interfacial area), heat and mass transfer measurements [ 33–44 ].
Typical examples from the literature include water, glycerol
[ 34 , 35 , 43 ], sucrose [ 42 ], sugar syrup [ 36 ], silicon and mineral oil [ 36 ],
castor and rapeseed oil [ 38 ], aqueous polyvinyl pyrrolidon solution
[ 40 ], polyethylene glycol (PEG) [ 43 ] and Na
2
SO
4
solutions [ 37 ] as
Newtonian fluids. As non-Newtonian model fluids carboxyl methylcel-
lulose (CMC) solutions [ 35–39 , 42 ], aqueous solutions of kaolin and
polyoxpolyacrylamide (PAA) [ 36 ], Carbopol [ 42 ], Xanthan gum
[ 39 , 42 , 43 ], soluble starch [ 34 ], neoponite, wallpaper paint and lapo-
nite solutions [ 36 ] have been used. Due to the complexity of rheological
measurements, the choice of model media is often based on an in-
adequate characterization of the real cultivation broth. A direct com-
parison has seldomly been conducted and the choice of model media
was to some extent done arbitrary in the literature [ 28 ].
As pointed out by many authors [ 30 , 45 ], measuring the rheology of
filamentous cultivation broths is challenging due to their strong in-
homogeneity. Therefore, this study intends to investigate the suitability
of rheometer configurations and measurement methodologies to char-
acterize the rheological properties of three filamentous microorgan-
isms. In particular, the study deals with the fungus Aspergillus niger and
the bacteria Lentzea aerocolonigenes and Actinomadura namibiensis . Ad-
ditionally, the work aims at the rheological characterization of the
batch cultivations of these three microorganisms based on the Ostwald-
de Waele power law model, enriching the database of the reported
rheological characterization for such cultivation systems. For the bac-
terial cultures, rheological properties are determined for first time. The
rheology of A. niger broths has been investigated by several authors, but
partially contradictive findings are reported [ 7 ], requiring further sys-
tematic investigations. Flow curves were measured considering culti-
vation time, biomass concentration and morphological shape de-
scriptors as the key parameters influencing the rheology of the
cultivation broth. However, several studies did not consider all influ-
encing parameters leading to erroneous conclusions. Finally, as men-
tioned before, many technical applications in literature employ non-
biological model fluids instead of the original biological matter, but the
choice of model fluids is often based on inadequate rheological com-
parison with the cultivation broths. Therefore, the last aim of the cur-
rent study is to compare the biological rheological behavior with that of
the model fluids widely used in the literature. This way the suitability of
such model fluids to mimic the rheology of cultivation broths in further
technical applications is tested.
2. Materials and methods
2.1. Filamentous microorganisms and cultivation systems
In this study, the rheological behavior of three filamentous micro-
organisms was investigated: the eukaryotic fungus Aspergillus niger and
the prokaryotic actinomycetes Lentzea aerocolonigenes and
Actinomadura namibiensis .
Aspergillus niger SKAn 1015 is a recombinant strain carrying the suc1
(fructofuranosidase) gene leading to the production of fructofur-
anosidase under the control of the constitutive pkiA (pyruvate kinase)
promoter [ 46 ]. Cultivations were conducted in a stirred tank bioreactor
(Applikon, Schiedam, The Netherlands) with 2.2 L filling volume. The
tank was equipped with three baffles and two six-bladed Rushton tur-
bine impellers (diameter 45 mm and impeller spacing 70 mm). In-
oculation and cultivation were carried out as described by Wucherp-
fenning et al. [ 46 ]. An aeration rate of 1 L/min, an agitation speed of
300 1/min, a temperature of 37 °C and a pH value of 5.0 were set as
operational conditions. The mycelial morphology was adjusted mainly
by NaCl salt addition (with exception the A. niger AS4 cultivation) af-
fecting the osmolality of the system, as described by Wucherpfenning
[ 21 ], and presented in Table 1 .
Lentzea aerocolonigenes DSM 44217, purchased from the German
Collection of Microorganisms and Cell Cultures (DSMZ, Braunschweig,
Germany), was cultivated in 250 mL shaking flasks with four baffles and
50 mL filling volume. So far, a method to cultivate L. aerocolonigenes in
bioreactors successfully has not been established. The shaking flasks
with GYM-Medium (4 g/L glucose, 4 g/L yeast extract, 10 g/L malt
extract, pH = 7.2) were incubated in an orbital shaker (Certomat BS-1,
Sartorius, Göttingen, Germany) with a shaking diameter of 50 mm and
a shaking speed of 120 1/min at 28 °C in darkness. Pre-cultures were
inoculated with 1 mL of frozen culture (biomass of a two-day pre-cul-
ture frozen at −80 °C in 30 % glycerol) and incubated for two days. The
main-cultivation was then inoculated with 300 μL pre-culture and in-
cubated for 8 days at the same conditions as the main-culture.
The filamentous strain Actinomadura namibiensis DSM 6313 used in
this study was obtained from the German Collection of Microorganisms
and Cell Cultures (DSMZ, Braunschweig, Germany). The strain was
cultivated as described by Tesche et al. [ 47 ]. Main cultivations were
inoculated with 5 mL of the pre-culture and cultivated in 500 mL
shaking flasks without baffles and 100 mL working volume for 240 h at
30 °C and 180 1/min (50 mm orbital shaking diameter) in darkness.
2.2. Model fluids
Seven types of products were tested for their suitability as model
fluid media: Inverted sugar syrup (72.7/66, Grafschafter Krautfabrik
Josef Schmitz, Meckenheim, Germany), Glycerin (≥ 98 % water free,
Carl Roth, Karlsruhe, Germany), Luviskol ® K90 Powder (BASF,
Ludwigshafen, Germany) (all Newtonian flow behavior), Polyethylene
oxide-PEG (molecular weight (MW) > 5 000 kDa, Alfa Aesar, Ward
Hill, USA), Xanthan gum (Food grade E415, Fufeng., Shandong, PRC),
Carboxymethyl cellulose solution (CMC Blanose, Ashland, Wilmington,
USA) (all shear thinning flow behavior). Aqueous solutions of the
abovementioned substances were prepared using ultrapure water
(electrical conductivity κ = 0.055 μS/cm, PURELAB flex 2 system,
ELGA Labwater, Celle, Germany). All solutions were prepared on a
weight basis ( Table 1 ) and used within one day after preparation.
2.3. Rheological Measurements
Due to the strong inhomogeneity of the cultivation broths, their
rheological characterization cannot be considered trivial. The main
challenges concern sample preparation and handling, settling or slip
conditions as well as sample alteration or damage during the mea-
surement [ 32 , 48 ]. Therefore, the first goal of this work was to
C. Bliatsiou, et al. Biochemical Engineering Journal 163 (2020) 107746
4

determine appropriate measuring configurations and methods for the
rheological analysis of different filamentous cultivation broths.
Cultivation broths were analyzed using a tempered rotational rhe-
ometer with air bearing (Kinexus lab+, Malvern Panalytical, Kassel,
Germany) and different measuring geometries: a plate-plate, a cone-
plate and an impeller configuration with a four-paddle. For each bio-
logical sample, a measuring geometry as well as the shear rate range for
accurate measurements were selected, depending strongly on the broth
viscosity, biomass concentration and most importantly the morpholo-
gical state of the biological matter, as it will be analytically discussed in
Chap. 3.1. To select the most appropriate measuring geometry for each
examined microorganism and sample, transparent acrylic glass dupli-
cates of the lower plate of the plate-plate configuration and the cy-
lindrical vessel of the impeller configuration were built. These trans-
parent set-ups ensured visual accessibility during the measurement. The
pellet suspension stage, any possible pellet settling and sample altera-
tion during the measurement were recorded by taking images of the
transparent set-ups with a digital camera. In the case of the transparent
plate-plate system, images were taken through an inclined mirror po-
sitioned below the plate. Images of the transparent vessel for the im-
peller set-up were taken from one side of the vessel.
Dimensions of the rheometer geometries are summarized in Table 2 .
The impeller rheometer configuration was pre-calibrated by the man-
ufacturer with fluids of known viscosity using the Metzner-Otto ap-
proach as applied in several works [ 12 ]. This set-up requires sig-
nificantly larger probe volume (32 mL) compared to the other
measuring systems (< 5 mL). For the impeller system, measurements
were performed in six logarithmic steps per decade from high to low
shear rates to obtain a well-mixed suspension. All other measuring
protocols included increasing shear rates. The measuring range of shear
rates varied depending on the sample properties as it will be analyti-
cally discussed in Chap. 3.1.
Treatment of the sample before measurement was shown to be
important [ 5 ]. Samples were withdrawn during the cultivation and
stored temporarily in a heat-bath at the optimum temperature for each
microorganism. Finally, they were poured on the measuring device and
directly measured. Measurements were carried out at ideal cultivation
temperature of the microorganisms which was 37 °C for A. niger , 28 °C
for L. aerocolonigenes and 30 °C for A. namibiensis, respectively. All
measurements were performed in duplicate.
To determine the dependency of the rheological behavior on bio-
mass concentration without taking into consideration any morpholo-
gical variations, selectively chosen broth samples were either diluted
with defined volumes of the respective cultivation medium or con-
centrated by removing an amount of the supernatant fluid to create
samples in a wide range of biomass concentration while the mor-
phology remained constant from sample to sample.
Rheological behavior of the model fluids was investigated using a
tempered cone-plate rheometer with air bearing (MCR 302, Anton Paar,
Graz, Austria). Measurements were carried out at 20
±

0.1 °C in the
range of shear rate = 0.01 - 5000 s

−1
.
2.4. Quantification of biomass dry weight concentration and cell
morphology
The biomass dry weight concentration ( BDW ) was determined
gravimetrically. Therefore, a pre-defined volume of the cultivation
broth was filtered through pre-dried and pre-weighted (24 h at 105 °C)
filter discs (Filter Discs Grade 389, Sartorius, Göttingen, Germany),
rinsed with deionized water and dried for 48 h at 105 °C. The sample
was then re-weighted. BDW was finally calculated as the dry weight of
the biomass in the sample volume. All measurements were performed in
duplicate.
The culture cell morphology was analyzed under an EVOS xl mi-
croscope with 4.5 μm/pixel (Thermo Fisher Scientific, Waltham, USA).
The bio-suspension sample was diluted in a petri dish to allow in-
dividual pellets to be examined. For each sample at least 100 pellets or
mycelial clumps were analyzed. Image analysis was conducted using
ImageJ 1.52 h (National Institute of Health, Bethesda, USA). The
images were first converted to 8-bit grayscale format and then binarized
using the thresholding tools in ImageJ. The particle analysis tool was
applied to calculate the desired morphological parameters of each
pellet or mycelial structure. Objects exceeding the edge of an image
were not considered for evaluation as they could not be analyzed as full
objects. Different morphological parameters can be obtained with this
method, however, only the maximum Feret diameter, defined as the
maximal distance between two points along the selection boundary of
the analyzed object, was considered for further analysis in the study.
The average bio-agglomerares diameter d
av
for each sample was cal-
culated as the arithmetic mean of the maximum Feret diameter of every
pellet.
3. Results and discussion
3.1. Measuring method for the rheological characterization of filamentous
cultivation broths
The first goal of this experimental work is to establish appropriate
measuring methodologies for the rheological analysis of different fila-
mentous cultivation broths. Three rheometer set-ups were tested for
their suitability: a cone-plate, a plate-plate and an impeller configura-
tion. The main challenges concern the avoidance of phase separation,
pellet settling (sustaining the particulate matter uniformly distributed)
and morphological alterations/damages of the bioagglomerates during
the measurement. The investigation of three strains with different
rheometers allows a comparative understanding of the influence of the
broth viscosity, biomass concentraion, pellet size and cell morphology
on the applicability of a measuring geometry and the respective suitable
shear rate measuring range.
3.1.1. Aspergillus niger
Samples of the cultivation AS2 of mycelial grown A. niger were in-
itially measured with the plate-plate configuration by adjusting the gap
size between the plates to 1 mm and 1.5 mm. Fig. 1 shows microscopic
images of A. niger before (Fig. 1A) and after (Fig. 1B) a measurement
with the plate-plate configuration (1.5 mm gap size) and shear rates up
to 500 s
−1
. Severe sample damage and aggregation occurred as proven
by the microscopic images (Fig. 1B) as well as by the visual observation
of the lower plate (Fig. 1C) after the measurement. Biomass aggregates
were formed, rolling into filaments, which were separated from the
media while shearing. Malouf [ 32 ] has referred extensively to such
morphological alterations, where the solid phase builds aggregates,
Table 2
Dimensions of used rheometer configurations.
Rheometer Measuring system Diameter d [mm] Gap size s [mm] Angle
[°] Impeller height h [mm]
Cultivation broths
Kinexus lab+, Malvern Panalytical Plate-Plate 60 (upper plate)/65 (lower plate) 1…1.5 [-] –
Cone-Plate 60 (upper plate)/65 (lower plate) 0.03 1 –
Impeller 25 Cup diameter: 27.5 [-] [-] 60
Model fluids
MCR 301, Anton Paar Cone-Plate 59.978 0.117 1.008 –
C. Bliatsiou, et al.

Biochemical Engineering Journal 163 (2020) 107746
5

when the parallel plate geometry was used for samples of Trichoderma
reesei . Rheological measurements for the case of the pelleted cultiva-
tions of A. niger (AS0, AS1) were not conducted with the plate-plate
system because the pellet size can reach several millimeters (> 2 mm)
and this falls far below the required minimum ratio of gap size to pellet
diameter of ten given by Barnes [ 49 ] and Brummer [ 50 ]. Consequently,
the plate-plate configuration was found to be unsuitable for rheological
measurements of the fungal cultivation broth of A. niger independent of
its cell morphology, confirming the reported findings from previous
studies [ 32 , 48 ]. Similarly, the cone-plate configuration with an even
smaller gap size also failed to give reliable rheological data for the
cultivation broth of A. niger confirming again the literature findings
[ 32 , 48 ].
Using the impeller/vane configuration, no sample damage or al-
teration occurred. This was confirmed by microscopic analysis of the
sheared sample after measurement. A change in mean bio-agglomerate
diameter of less than five percent was observed in general ( Fig. 4 ). All
in all, the impeller rheometer was considered to be appropriate for the
rheological characterization of the fungal broth of A. niger showing a
good compatibility for both the mycelium and the pelleted growth,
justifying the common use of this set-up in literature. Nevertheless, a
main drawback of the method is the need of large sample volumes.
Significantly larger sample volumes (32 mL for the used vessel) were
required compared to the other measuring systems (< 5 mL for cone- or
plate-plate). Another important challenge of the method is the necessity
to keep the biomass in suspension, ensuring a homogeneous sample
distribution during the measurement, no phase separation and no for-
mation of less dense layers on the walls of the cylindrical vessel. These
points are especially crucial for low viscous broths and they have not
been considered sufficiently in literature yet. To face this, in general,
measurements were carried out from high to low shear rates to provide
a homogeneous pellet distribution. Additionally, to further evaluate the
suitability of the method and to identify a reasonable shear rate range
for the analysis of the flow curves, measurements were carried out with
a transparent vessel as described in Chap. 2.3. The images taken by the
camera were analyzed to determine the minimum shear rate
min

needed to fully suspend the microorganisms. Fig. 2 shows the respective
measurements with the transparent vessel for the pelleted cultivation
Fig. 1. Sample of the AS2 broth of A. niger at cultivation time t = 27.5 h, with BDW = 1.79 g/L and d

av
= 0.74 mm: (A) Microscopic image of the original sample, (B)
microscopic image of the sheared sample after the measurement with plate-plate rheometer, (C) image of the lower plate of the plate-plate configuration after the
measurement. A. niger is clearly aggregated into filaments.
Fig. 2. Evolution of the suspension state of A. niger broths with different cell morphology and biomass concentrations in the vane system at different shear rates: (A)
Cultivation AS0 (without NaCl addition), inoculum concentration 10

6
spores/mL, (B) Cultivation AS1 (without NaCl addition), inoculum concentration 3.9∙10
5
spores/mL, (C) Cultivation AS4, pH-shift.
C. Bliatsiou, et al. Biochemical Engineering Journal 163 (2020) 107746
6

AS0 and AS1 and the mycelial cultivation AS4 of A. niger . Comparative
measurements with the same sample in both original set-up and acrylic
glass duplicate showed a good agreement.
In Fig. 2 the images depict the suspension state under exemplary
shear rates for samples withdrawn at different times of the batch pro-
cesses. For the pelleted broth of A. niger AS0 ( Fig. 2 A), at cultivation
time t = 20.5 h a minimum shear rate of
min

= 50 s

−1
was needed to
completely suspend the biological matter. At later stages of the batch
cultivation ( t = 53.5 h), a homogeneous suspension could be reached at
lower shear rates allowing measurements even at
min

= 0.1 s

−1
. Al-
though pellet size increases over time implying a faster sedimentation
velocity for a single pellet, the growth of biomass increases the viscosity
of the cultivation broth, as it will be shown in detail in Chap. 3.2.1,
resulting in an overall decrease in sedimentation velocity. For the pel-
leted cultivation of A. niger AS1 the bio-agglomerates could not be
suspended within the applied range of shear rates even at the latest
stages of the process. Fig. 2 B shows the respective suspension state
during the measurement with the transparent vessel at t = 53.5 h. In
this case, the pellets were significantly larger in comparison to the
cultivation AS0 with an average size of d
av,AS1, t =53.5h
= 2.86 mm (in-
stead of d
av,AS0, t =53.5h
= 1.95 mm). At the same time, the pellet
structure was very compact with shorter exposed hyphae surrounding
the bio-agglomerate main core and the biomass concentration was
significantly lower with BDW
AS1, t =53.5h
= 1.05 g/L instead of
BDW
AS0, t =53.5h
= 4.56 g/L. Under these conditions, a rapid settling of
the pellets was observed independent of the agitation intensity and
thus, no suitable measuring method could be identified for the rheo-
logical characterization of the A. niger AS1 cultivation. Lin et al. [ 51 ]
compared the settling velocity of A. niger pellets with different pellet
surface structure but similar pellet diameter. They reported three times
higher sedimentation velocities for pellets with compact shape com-
pared to pellets with unstructured and irregular periphery. This con-
firms that the pellet size as well as other morphological surface char-
acteristics are significant when defining the appropriate measuring
range for rheological measurements. Fig. 2 C shows the suspension state
during measurements of the mycelial broth of A. niger AS4. The images
depict the suspension state under exemplary shear rates for samples
with the same bio-agglomerate morphology (mycelial growth,
d
av,AS4, t =47h
= 0.53 mm), but different biomass concentrations. By vi-
sual observation at the lower biomass level of BDW = 1.57 g/L ag-
glomerates accumulated in two distinct zones even at high shear rates
of
= 200 s

−1
. As the biomass concentration increases, the minimum
shear rate
min

for a complete dispersion decreased significantly. For the
highest biomass concentration of BDW = 7.37 g/L, mycelia appear to
be homogeneously distributed throughout the measurement even at a
low shear rate of

min

= 0.25 s

−1
. Thus, the biomass concentration is
another significant parameter when defining the appropriate measuring
range for rheological measurements, since it affects dramatically the
pellet-to-pellet interaction and so the particle settling.
3.1.2. Lentzea aerocolonigenes
Low reproducibility was recorded using the plate-plate configura-
tion for the broths of L. aerocolonigenes (LA) as depicted in Fig. 3 A. By
conducting measurements on the transparent plate, phase separation
was observed during the measurement ( Fig. 3 B and C). In Fig. 3 A the
deviation between the flow curve obtained with the acrylic glass plate
and the ones obtained with the original Kinexus lab + plate might be
the result of the broth inhomogeneity during the shearing process. Due
to the difficulty to ensure a reliable usage of the plate-plate rheometer
configuration for the broth characterization, further measurements
were conducted with the impeller/vane system, which showed re-
producible results. No morphological damage occurred during the
measurement as confirmed by the microscopic analysis of the sheared
sample ( Fig. 4 ). In this case, the transparent vane vessel was used again
to determine the minimum shear rate
min

. Comparative measurements
with the same sample in both the original vessel and the acrylic glass
duplicate showed a good agreement (compare Fig. 3 A). An increase in
biomass concentration again resulted in lower

min

for a complete dis-
persion, as a result of the higher viscosity of the broth at higher BDW .
The impact of morphology on

min

was not examined for L. aero-
colonigenes . This is because, due to the low broth volume (50 mL in
shaking flask), samples for the rheological characterization were taken
only at the end of the cultivation when the microorganism is in the
production phase and thus all samples of L. aerocolonigenes had the
same morphological state.
3.1.3. Actinomadura namibiensis
For the cultivation of A. namibiensis (AC), investigations with the
transparent vane geometry showed a rapid settling of the bio-
Fig. 3. (A) Rheograms of L. aerocolonigenes LA broth measured with different configurations. Top view image of broth sample on the transparent plate (B) before and
(C) during the measurement.
Fig. 4. Particle size cumulative distributions for samples of A. niger - AS2 ( t =
57.5 h, BDW = 5.83 g/L), L. aerocolonigenes – LA ( t = 192 h, BDW = 3.73 g/L)
and A. namibiensis - AC ( t = 96 h, BDW = 8.22 g/L) cultivations before and
after measurements with the system specific proper rheometers.

C. Bliatsiou, et al. Biochemical Engineering Journal 163 (2020) 107746
7

agglomerates of broth samples taken within the first two to five days of
the cultivation, resulting in low reproducibility. Only for the final stages
of the cultivation the sedimentation velocity was low enough allowing
the pellets to be suspended. Unlike A. niger and L. aerocolonigenes , the
average pellet diameter of A. namibiensis decreases with time. This
bacterium develops a highly heterogenous mycelial cell morphology.
During the cultivation time the amount of pellets decreases, whereas
the fraction of freely dispersed mycelia and loose aggregates increases
as reported by Tesche et al. [ 47 ]. The change in cell morphology and
the increase in biomass concentration explain the significant lower se-
dimentation velocities after approximately five days of cultivation.
Nevertheless, measurements with the vane geometry even at these later
stages were treated critically. The impeller rheometer was not further
used for this biological system.
On the contrary, due to the smaller average pellet size of A. nami-
biensis ( d
av
= 0.28−0.44 mm) compared to A. niger ( d
av
up to 3 mm
depending on the cultivation batch), the plate-plate configuration was
considered a reasonable option. Although the ratio of gap size to pellet
diameter falls below the recommended minimum value of ten [ 49 , 50 ],
the flow curves showed good reproducibility. This should be attributed
additionally to the structural properties (micro-morphology) of this
bacterium, that apparently ensure high stability of the formed pellets.
Using the acrylic glass duplicate of the plate-plate configuration a
homogeneous distribution of the biomass was seen. Since stainless steel
and acrylic glass have different surface properties, the adhesion in the
replica system may have differed from the original plate-plate system.
However, the obtained values of μ
app
were very similar for both original
and acrylic plate-plate systems. Therefore, it is assumed that the dis-
tribution of the biomass is similar in the original and the replicated
plate-plate system. To assess, if bio-agglomerates changed in size or
structure, image analysis was performed before and after a measure-
ment. Fig. 4 shows the pellet-size distribution before and after a mea-
surement with the plate-plate rheometer. No significant change was
observed. Based on the findings, the plate-plate rheometer was con-
sidered a suitable measurement system for A. namibiensis for the entire
cultivation period.
3.1.4. Conclusions: measuring methods
The impeller rheometer allows a precise rheological characteriza-
tion of A. niger and L. aerocolonigenes cultivation broths providing re-
sults with an acceptable accuracy and reproducibility in most cases. It
was proven that this set-up prevents alteration of the morphology of the
biological sample.
Based on the measurements with the transparent vane vessel, care
must be taken when choosing the appropriate measuring range of shear
rate. The impeller rheometer can offer reliable data only for a limited
range of shear rate that depends strongly on the cell morphology and
biomass concentration of the broth. In most cases, the minimum shear
rate should be as high as
min

10 s

−1
to avoid sedimentation and
phase separation. At the same time, the measurements can be con-
ducted as long as turbulent effects are avoided for the Metzner-Otto
concept to be valid. Thus, the upper shear rate limit depends on the
viscosity of the sample (see Fig. A1 ). The more viscous the broth, the
higher the shear rates that can be applied. By measuring model fluids of
known rheology, Re < 150 was found as a suitable criterion to avoid
turbulent effects. Therefore, in the case of the biological samples, the
maximum shear rate
max

for the evaluation was chosen according to
this Reynolds number limit for each sample. Even though the laminar
flow regime ranges up to Re 10, the Metzner-Otto concept can still be
valid in the transitional regime and therefore at Re < 150 [ 52 ]. Sedi-
mentation cannot be avoided for low viscous samples or large particles
with compact and smooth surface structure. Therefore, there are lim-
itations when measuring low viscosities associated especially with the
first samples of a cultivation. The problem was already mentioned in
the literature by Malouf [ 32 ], who suggested the usage of a concentric
cylinder for the measurements of the early obtained broth samples. In
preliminary tests the use of such system showed low reproducibility and
significant morphological alteration for A. niger , mainly due to the
significant larger bio-agglomerate sizes in comparison to the cultiva-
tions of Malouf [ 32 ] with d

av,max
100 μm.
The impeller rheometer did not seem appropriate for a precise
rheological characterization of the A. namibiensis cultivation broth. On
the contrary, the plate-plate rheometer configuration provides results
with an acceptable accuracy and reproducibility. The use of this system
should be interpreted as the result of both macro- (size, shape) and
micro- (structure) morphology of the bacteria that ensure the avoidance
of pellet disintegration/alteration during shearing.
3.2. Rheological characterization of filamentous cultivation broths
For all samples of the different cultivations the measuring data of
shear stress as a function of the shear rate were plotted in double
logarithmic plots and fitted to the power law model. Despite the fact
that the power law cannot describe yield stresses, this model was
chosen for all following results, because of its simplicity and since
measurements at very low shear rates (to determine yield stresses)
would be considered at least critical if not unreliable as described in
Chap. 3.1. For shear thinning samples, the flow consistency factor K and
the flow behavior index n of the power law model were determined by
linear regression. For each experimental point, at least two measure-
ments (each with new sample) were carried out to ensure the re-
producibility of the recorded flow curves (exemplarily see Fig. A1 ). The
error bars in the following graphs depict the deviations between mul-
tiple samples.
3.2.1. Cultivation of Aspergillus niger
A. niger was cultivated under controlled level of osmolality (culti-
vation AS0, AS1, AS2, AS3) or pH (AS4) to obtain a desired fungal
morphology (see Table 1 ). Based on the work of Wucherpfennig et al.
[ 46 ], in this study the osmolality level at the various cultivations was
increased from 400 mosmol/kg (without NaCl addition) to 1500
mosmol/kg ( c

NaCl
= 32.18 g/L) and further to 2500 mosmol/kg ( c
NaCl
= 64.43 g/L), resulting in a change of the fungal cell morphology from
the pelleted to a mycelial form and decreasing bio-agglomerate sizes.
The obtained morphological growth was in a good agreement with that
described in Wucherpfennig et al. [ 46 ].
The dynamic viscosity of the Newtonian culture medium as mea-
sured at the beginning of the cultivation was μ = 0.844 mPa·s at 37 °C.
At various time points during the cultivation, samples of the broth were
filtered to separate the biomass and the supernatant viscosity was
measured. It was proven that during the cultivation the viscosity of the
continuous phase does not change, but remains at water-viscosity level,
meaning that the rheological changes described in the following para-
graphs are caused exclusively by the presence of bio-agglomerates and
not by viscous metabolites released to the medium.
3.2.1.1. Broth rheology during the cultivation . At an osmolality of 400
mosmol/kg (cultivation AS0), A. niger grew in large spherical pellets.
The growth phase of the cultivation lasted approximately 55 h before
reaching the stationary and later the lethal phase. In this time period,
the evaluation of the rheological measurements succeeded from a shear
rate of
min

= 1 s

−1
and at the latest stage of the cultivation ( t ≥ 50 h)
from a shear rate of
min

= 0.1 s

-1
to avoid sedimentation of the pellets.
For the pelleted cultivation broth AS0, at BDW ≤ 0.5 g/L the low
viscosities did not allow to obtain reasonable measuring data. At this
initial process stage, the flow behavior index was considered to be near
n ∼ 1, dominated by the Newtonian behavior of the cultivation
medium. At BDW > 0.5 g/L the broth showed shear thinning
behavior and higher viscosities. Once the stationary phase was
reached, basically no further change of the flow curves was recorded.
Even though in literature it is often claimed that pelleted broths present
C. Bliatsiou, et al. Biochemical Engineering Journal 163 (2020) 107746
8

Newtonian rheological behavior and low viscosities [ 20 , 53 ], the
findings of the present work do not confirm this statement, which
seems to be rather generalized. In the past, other authors have also
reported non-Newtonian rheology of pelleted broths [ 11 , 27 , 28 , 54–56 ].
In Fig. 5 A the development of the rheological parameters, the flow
consistency factor K and the flow behavior index n , the BDW and the
morphological characteristic (average pellet diameter d
av
) are plotted
as a function of cultivation time for the pelleted AS0 broth. As a result
of the increase in BDW, the growth of the pellets and thus the increasing
pellet-to-pellet interactions [ 57 ], an increase in K is observed. For
pelleted AS0 broth (inoculum concentration 10
6
spores/mL) in the
examined cultivation period, i.e., t = 20–53.5 h and level of biomass,
i.e., BDW = 0.58–4.69 g/L, the flow consistency factor varied in a
range of K = 0.01−0.65 Pa∙s
n
. The flow behavior index after t =20 h
appears to reach a plateau, i.e., n = 0.24 ± 0.03, without being
directly influenced by the evolution of the other system properties.
The cultivation A. niger AS1 (not shown) had again an osmolality of
400 mosmol/kg (no NaCl addition), but the inoculum spore con-
centration was decreased compared to AS0 cultivation to 3.9∙10
5
spores/mL. A pelleted growth was obtained again, but the pellets were
significantly larger in comparison to AS0 cultivation (compare
Fig. 2 A,B, average particle diameter d
av
up to 2.9 mm at AS1, in con-
trast to d
av
< 2 mm at AS0), the pellet structure was more compact with
shorter hyphal length and the BDW was approximately four times lower
than for AS0. The low viscosities and the fast sedimentation of the
pellets made any quantified rheological characterization of the broth
impossible as previously described in Chap. 3.1.
At an osmolality of 1500 mosmol/kg (cultivation AS2), A. niger grew
in the form of freely dispersed mycelia. In this case, the cultivation had
a total duration of 107 h ( Fig. 5 B) and the rheological data used for the
evaluation included shear rates higher than
min

= 10 s

−1
. The vane
measuring system could be used successfully only at BDW > 2 g/L. The
rheological data reported below this level should be treated with care,
since high deviations were observed during the measurements, due to
the low viscosities of the samples. At higher biomass concentrations, the
broth exhibited a pseudoplastic behavior, which was attributed to hy-
phal entanglement during the growth [ 54 ]. Similar to the pelleted
cultivation AS0, again in this case an increase in K was reported during
the cultivation. At approximately 80 h the cultivation reached the sta-
tionary stage. It is remarkable that the growth period for the A. niger
AS2 cultivation lasted longer in comparison to the A. niger AS0 culti-
vation, since the mycelial forms with low compactness are more active
and suffer less from mass transport limitations. After 80 h of cultivation,
a decrease in K was observed. This decrease indicates the lethal phase of
the cultivation and it was observed for all cultivations of Table 1 . The
decreased viscosities at these stages seem to be the result of hyphae
fragmentation and lysis. The structural changes of the cells at the late
cultivation stages make them less resistant to shear forces. At the same
time, the formation of small spores, which typically appear at the
branched hyphae and have much lower viscous properties than the fi-
laments, results in a drop in the flow consistency factors K [ 14 , 32 , 57 ].
Again, no clear correlation of n with the evolution of the BDW or the
hyphae size is reported. For the examined cultivation period and level
of biomass, i.e., t = 43–107 h and BDW = 3.7–6.3 g/L, the flow con-
sistency factors varied in a range of K = 0.4–2 Pa∙s
n
and the flow be-
havior index reached a plateau of n = 0.27 ± 0.07 for the broth of the
freely dispersed mycelia ( Fig. 5 B).
At an osmolality of 2500 mosmol/kg (cultivation AS3, inoculum
concentration 10
6
spores/mL, not shown), A. niger grew in small
elongated clumps. In this case, the system reached the stationary stage
after 80 h of cultivation. The maximum level of biomass determined
was approximately BDW ∼ 3.0 g/L and A. niger built compact ag-
glomerates of average sizes in the range of d
av
∼ 0.4−0.7 mm. For this
cultivation broth the use of the vane geometry was significantly limited
due to apparent pellet sedimentation. Even at the latest stages of the
cultivation, the flow consistency factors lay at the level of K ∼ 0.05
Pa∙s
n
. At these low broth viscosities, the rheological data cannot be used
reliably. In their study with Aspergillus oryzae, Müller et al. [ 58 ] con-
nected also small size clumps and low biomass concentrations to low
viscosities. A plausible explanation could be that small clumps possess
shorter hyphae that cannot contribute significantly to pellet-to-pellet
interactions and so the viscosity remains at low levels independent of
the BDW level.
To understand better the rheological behavior of small elongated
clumps, the cultivation A. niger AS4 (without NaCl addition, inoculum
concentration 10
6
spores/mL) was carried out. In this case, the pH was
used to adjust the agglomerate morphology at the desired level (pH-
shift from 3 to 5 from cultivation time 24 h–29 h). The advantage of this
cultivation in comparison to the AS3 cultivation is that the biomass
grew up to BDW = 4.98 ± 0.31 g/L at t = 47 h. Thus, even though at
a biomass concentration of up to 2.5 g/L similar problems were faced
during the measurements as for the AS3 samples, at higher BDW
rheological data were reliably obtained and they will be discussed in
the following section.
3.2.1.2. Influence of biomass on broth rheology . Fig. 6 depicts the
evolution of K and n as a function of BDW under identical
morphological state; with d
av
=1.95 mm for the pelleted broth AS0
Fig. 5. Evolution of the broth properties during the cultivation of A. niger for (A) the pelleted AS0 broth and (B) the mycelial AS2 broth.

C. Bliatsiou, et al. Biochemical Engineering Journal 163 (2020) 107746
9

( Fig. 6 A), d
av
=1.33 mm for the mycelial broth AS2 ( Fig. 6 B) and d
av
=0.54 mm for the mycelial broth AS4 ( Fig. 6 C). For these
measurements an original sample taken from the stirred tank
bioreactor during the cultivation was diluted (with cultivation
medium at different ratios) or/and concentrated (removal of
supernatant after biomass sedimented), to obtain broth samples in a
wide range of BDW .
Here, the clear significant dependence of K on BDW is shown. With
increasing BDW , the K value rises. When trying to build a mathematical
relationship with a power law approach to describe the dependence of
K on BDW - in the form of K = C
K
·
BDW

- for each cultivation, an ex-
ponent in the range of = 1.65–1.8 was determined. However, such
simple mathematical relationships in the form of a power law approach
should be treated carefully, because the amount of available data may
not be sufficient for a statistically safe mathematical modeling. Also, as
pointed out by Allen and Robinson [ 13 ], such correlations should be
treated as system specific and not as “universally” applicable models.
Thus, even though the exponent found in this work seems to be in
general agreement to similar approaches in literature; Albaek et al. [ 59 ]
reported = 1.5–1.56 for A. oryzae , Olsvik et al. [ 25 ] reported = 1.3
for A. niger , Gabelle et al. [ 41 ] reported = 1.5 for T. reesei , and
Pamboukian and Facciotti [ 55 ] reported = 2.03 for Streptomyces
olindensis , a comparison with literature correlations is generally diffi-
cult because of the differences in cultivated strain, cultivation condi-
tions, cell morphologies and measuring methods.
No clear correlation between n and BDW was recorded for any cell
morphology of A. niger , with n being almost independent of the biomass
level. The average value of n for the measured range of biomass con-
centration was n = 0.29 ± 0.07 for the pelleted A. niger AS0 cultiva-
tion ( BDW = 0.58–4.56 g/L), n = 0.26 ± 0.06 ( BDW = 1.91–5.83 g/
L) for the mycelial-growing AS2 broth and n = 0.29 ± 0.07 ( BDW =
1.57–7.37 g/L) for the mycelial AS4 broth, respectively ( Fig. 6 A–C).
Various studies have reported mathematical modeling - usually also
in the form of the power law approach of n = C

n
·
BDW

- to correlate n
with BDW . However, other studies claim that such a relationship is not
so straightforward, as confirmed by the present work as well. In the
study of Riley et al. [ 20 ] for the tested cultivations of P. chrysogenum ,
the n value was principally stable with an average value of 0.35 for
BDW > 10 g/L with exceptions of the values measured at the earliest
and latest cultivation phases. Similar findings were reported earlier by
Pedersen et al. [ 28 ] for P. chrysogenum with n appearing to be ap-
proximately constant and independent of the BDW reaching a plateau at
0.4−0.5 at the examined biomass range. Müller et al. [ 58 ] also con-
cluded that in the cultivations of A. oryzae no correlation could be
found between n and BDW or the examined strains of Aspergillus , with
the recording range of n values to fluctuate strongly in a range of
0.03−0.38 under varying pH cultivation conditions and morphological
states of the different employed strains. The early work of Allen and
Robinson [ 24 ] stated that for the filamentous cultivation broths of A.
niger , P. chrysogenum and Streptomyces levoris , with increasing biomass
concentration n drops off rapidly from 1 to values between 0.2−0.4. In
agreement with that, in the work of Pollard et al. [ 7 ], the filamentous
Fig. 6. Impact of biomass concentration BDW on the flow consistency factor K and flow behavior index n for (A) the pelleted AS0 broth, (B) the mycelial AS2 broth,
(C) the mycelial AS4 broth of A. niger and (D) the pelleted LA broth of L. aerocolonigenes .

C. Bliatsiou, et al. Biochemical Engineering Journal 163 (2020) 107746
10

fungi Glarea lozoyensis also showed that n dropped rapidly at the initial
stages of cultivations to constant values in a range of 0.35−0.4. Pam-
boukian and Facciotti [ 55 ] characterized batch and fed-batch cultiva-
tions of S. olindensis and reported a sharp decrease in n after the initial
24 h of cultivation, which then remains relatively stable ( n ∼ 0.3),
while the cell morphology had hardly an influence on the n values.
Similar findings have been recorded by Gabelle et al. [ 41 ] for cultiva-
tions of T. reesei where n was found constant ( n = 0.4 ± 0.1) for
BDW > 7 g/L. Petersen et al. [ 12 ] claimed no correlation between n and
biomass concentration, batch time, agglomerates size or feed mode for
a pilot scale cultivation of A. oryzae . Finally, Olsvik et al. [ 25 ] in their
work with A. niger could not find any strong dependence of n on the
cultivation conditions.
3.2.1.3. Influence of macro-morphology on broth rheology . Throughout
the experimental study, it was observed that the effect of the bio-
agglomerate size on the broth viscosity was not as strong as the effect of
the biomass concentration and the effect of other morphological
characteristics (e.g., compactness, roughness, hyphal length
surrounding the pellet core). This point was first reported in Chap.
3.1 when a comparison was made between AS0 and AS1 cultivation
broth. Despite the bigger pellet size of the A. niger AS1 cultivation, the
lower biomass concentration, the high pellet compactness and the
lower pellet roughness made the cultivation broth less viscous than the
AS0 broth and led to faster pellet settling. In the current chapter, by
comparing Fig. 5 A and B (rheological characterization at different
points of time during the cultivation) with Fig. 6 A and B (rheological
characterization at specific cultivation point with concentrated/diluted
samples of the same bio-agglomerate morphology) respectively (see
also Fig. A2A), again it can be observed that the impact of the biomass
concentration on the viscosity of the system and thus the flow
consistency factor K is significantly stronger than the one of the bio-
agglomerate size. Figure A2A shows that with the same biomass
concentration the recorded K values are very similar independent of
the agglomerate diameter.
This allows merging the data of Fig. 5 and Fig. 6 (see Fig. A2 B) and
draw conclusions for the effect of other morphological factors (e.g.,
shape, compactness) on flow consistency factor K under consideration
of the same biomass concentration. As depicted in Fig. A2 B only a slight
difference in the flow consistency factor K was recorded between the
AS0 (pellets) and the AS2 (mycelia) cultivations. At the same biomass
concentration, independently of agglomerate size, the AS2 broth was
slightly more viscous (higher K values) than the AS0 broth. In both
broths, A. niger built agglomerates with fluffy, rough external surface,
which contributes to the pellet-to-pellet interactions and the increase in
viscosity. On the other side, the elongated compact clumps of the AS4
cultivation with the short hyphae hair (decreased roughness) had
clearly lower K and, therefore, lower viscosities in comparison to the
AS0 and AS2 cultivations. Thus, the external pellet surface properties
seem to be crucial for pellet-to-pellet interaction and so the viscosity of
the system. The inner core of the bio-agglomerate seems to affect in-
directly the viscosity of the system, by defining the biomass growth.
The pellets of the AS0 broth had a compact core, which limited the
oxygen and nutrient transport and finally led to an earlier fungi death
and lower total biomass levels. A. niger of the AS2 cultivation had a
highly loose inner structure leading to better mass transport, prolonged
growth and eventually higher biomass concentrations, at which sig-
nificantly higher viscosities were recorded. Olsvik et al. [ 25 ] in their
work with A. niger , after proving that the biomass is the most crucial
system variable affecting the flow consistency factor, they examined
systematically the influence of various morphological parameters, i.e.,
roughness, compactness, clump area and overall percentage of mycelial
in form of clumps, on K . It was reported that at the same biomass
concentration, the flow consistency factor was strongly correlated to
the clump roughness, to lesser extent to the compactness, while K was
not significantly interrelated with other morphological parameters.
This statement is in agreement with the findings of the current work.
Despite the distinct cell morphologies (diameter, shape, roughness,
compactness etc.) of A. niger , no significant difference in flow behavior
index n was recorded for the AS0, AS2 and AS4 cultivations as seen in
Fig. 6 . Indisputably, this result, referred previously also in other studies
such as the ones of Olsvik et al. [ 25 ], Pamboukian and Facciotti [ 55 ]
and Müller et al. [ 55 ], indicates the need to relate the degree of pseu-
doplasticity more to micro-morphology, the inner structural properties
and the inner stability of a specific strain and not to macro-morphology.
At the same time, the results of A. niger under salt exposure (AS2)
are in contrast to Wucherpfennig et al. [ 15 ]. The authors, who were the
first to rheologically investigate this specific strain of A. niger, reported
a strong influence of BDW on the flow behavior index n and distinguish
three different areas: shear-thickening behavior ( n > 1) for low BDW
(0.5–2.5 g/L), Newtonian behavior ( n = 1) for medium BDW
(2.5–4.0 g/L) and shear-thinning flow behavior ( n < 1) for high BDW
(4.0–9.0 g/L). A shear-thickening behavior seems very unusual and the
data generated in this work with n ≤ 1 for all cultivation times do not
confirm the findings of Wucherpfennig et al. [ 15 ]. Furthermore, the K
values determined between 0.07 and 0.75 Pa∙s
n
(compare Fig. 6 B, AS2,
addition of 32.18 g/L NaCl) are much more plausible than the pre-
viously determined K values [ 15 ], which are at levels far below the
viscosity of water. The deviations between the current work and that
one of Wucherpfennig et al. [ 15 ] clearly show how critically rheological
investigations of filamentous systems have to be evaluated and that a
great effort of research is still needed to improve the accuracy and re-
producibility of such investigations.
In literature authors have reported predictive correlations that de-
scribe the interrelation among broth rheology, biomass concentration
and cell morphology. For A. niger, such correlations describe the flow
consistency factor K as a function of the biomass concentration and the
pellet roughness [ 5 , 25 ] or the non-dimensional Morphology number
[ 15 ]. Difficulty arises to use these correlations, among others due to the
variety of methods and parameters used to describe the moprhological
characteristics of the bio-agglomerates.
For the prediction of such sophisticated dependencies though, the
concept of the effective volume fraction Φ
eff
might be of interest. This
concept considers the solid volume fraction and the liquid that can be
embedded between solid agglomerates. This amount of liquid is tem-
porarily immobilized in the agglomerates and it should therefore not be
considered in the liquid volume fraction [ 60 ]. The solid volume fraction
in this study is composed of biomass, which significantly influences the
viscosity of the cultivation broth. The effective solid volume fraction is
potentially even higher, especially in case of pellets or mycelia with a
rough and irregular surface, which tend to entangle. At high biomass
concnentrations, the potential for entaglement and embedded liquid
volume fraction is increased, since a lot more particle contacts are
possible. At lower biomass concentrations, however, the interactions
between pellets or mycelia are lower due to larger distances in between.
Such behavior was described for the viscosity of suspensions by Nguyen
et al. [ 61 ]. Accessing the effective volume fraction is not easy for such
complex biological systems and thus it was not investigated in detail in
this study. Nevertheless, this concept can be promising for further re-
search.
3.2.2. Cultivation of Lentzea aerocologineses
As described in Chap. 3.1 the vane geometry was used for this
biological system. The batch cultivation of L. aerocolonigenes lasted in
total 8 days. However, due to the limited broth volume (50 mL in
shaking flask), samples for the rheological characterization were taken
only at the end of the cultivation, since the microorganism is in the
production phase at this point. The rheology of the LA cultivation broth
at t =192 h was determined. Additionally, the influence of the biomass
on the broth rheology was investigated by following the same metho-
dology of diluting and concentrating the originally obtained sample as
in the case of the A. niger cultivations, thus, creating samples with the
C. Bliatsiou, et al. Biochemical Engineering Journal 163 (2020) 107746
11

same cell morphology but varying solid content. The dynamic viscosity
of the Newtonian culture medium was measured to be μ = 0.93 mPa·s at
30 °C. At t =192 h, a sample of the broth was filtered to separate the
biomass and the supernatant viscosity was determined resulting in
μ = 0.96 mPa·s, indicating that the increased viscosities and the pseu-
doplasticity of the broth depicted in Fig. 6 D was entirely due to the
biomass growth and not the aqueous culture medium or extracellular
metabolites.
The rheological data used for fitting included shear rates higher
than
min

=10 s

−1
for BDW < 1.5 g/L. Above this biomass level the
minimum shear rate taken into consideration was
min

=1 s

−1
. As
shown in Fig. 6 D the cultivation broth is characterized by low viscosity,
since significantly small values of K were determined for all samples.
When BDW < 1.0 g/L, the broth has Newtonian behavior. Above this
biomass level, shear thinning properties were recorded. An increase of
BDW led to a clear increase of the K factor. At the same time, BDW had
no effect on n , similarly to what was reported in Chap. 3.2.1 for the
broths of A. niger . The average value of n for the measured range of
BDW was n = 0.41 ± 0.01 ( BDW = 1.29–4.29 g/L). In comparison to
A. niger , L. aerocologineses shows a less pronounced shear thinning be-
havior (higher n values) and lower apparent viscosity (lower K values)
at the same BDW . This might be explained by the difference in cell
morphology. The surface structure of A. niger has higher hyphal gra-
dients in the outer pellet layer compared to L. aerocolonigenes pellets
with a higher circularity/compactness surface structure.
3.2.3. Cultivation of Actinomadura namibiensis
As described in Chap. 3.1 the plate-plate geometry was used for A.
namibiensis . The batch cultivation of A. namibiensis lasted in total 10
days and this bacterial strain develops a highly heterogeneous mycelial
morphology. A transition from pellet to dispersed morphology takes
place during the cultivation, with the pellet to disintegrate into looser
fragments. Fig. 7 shows the rheological properties, the BDW con-
centration and the particle size d
av
during the AC cultivation. The
cultivation broth develops a shear thinning behavior. For the first 120 h
of the cultivation, as the biomass increased, the viscosity and thus the
flow consistency factor K increased. At this point in time, the pellet
structure started getting loose, the pellet size decreased and the amount
of freely dispersed mycelial increased. From this point on, no further
significant change in the rheology of the system was observed and the
measured flow curves were almost identical with
K = 0.39 ± 0.02 Pa∙s
n
and n = 0.40 ± 0.02 for t = 144–220 h.
3.3. Use of model fluid systems to simulate the biological rheology
Apart from the rheological characterization of the biological fila-
mentous systems, this work aims to find model fluid substances that can
mimic the rheology of the cultivation broth and can be further used for
various fluid dynamic and mass transfer investigations in the respective
bioreactors. Therefore, various substances were tested as suitable model
fluids (compare Table 1 ). The rheograms were recorded for all model
fluids. Fig. 8 depicts exemplarily flow curves of the non-Newtonian
model fluids (Xanthan gum, CMC and PEG solutions, Fig. 8 C) that were
examined, as well as the viscosity of the Newtonian fluids (inverted
sugar syrup, glycerin and Luviskol ® K90 solutions, Fig. 8 D) as a function
of the per weight concentration (% w/w) of each model substance in
aqueous solutions. It should be underlined that some model fluids de-
pending on the producer could exhibit deviating rheological properties.
The results shown in this work are valid for the specific substances as
described in Chap. 2.2.
Since all filamentous biological systems exhibit shear-thinning be-
havior, their flow curves were compared with those of the non-
Newtonian, shear thinning solutions: Xanthan gum, CMC and PEG so-
lutions. However, both CMC solutions ( Fig. 8 B) and PEG solutions
( Fig. 8 C) show lower pseudoplasticity in comparison to Xanthan solu-
tions ( Fig. 8 A) as well as the cultivation samples. Therefore, these
substances were not further considered as potential model fluids of the
examined filamentous cultivations.
Fig. 9 depicts exemplary flow curves of Xanthan gum solutions and
the cultivation broths of A. niger AS0 ( Fig. 9 A), AS2 ( Fig. 9 B) and AS4
( Fig. 9 C), L. aerocolonigenes (LA) ( Fig. 9 D) and A. namibiensis (AC)
( Fig. 9 E).
In general, the shear thinning behavior of A. niger can be simulated
satisfactorily by Xanthan gum solutions. Despite the fact that A. niger
exhibits a slightly stronger pseudoplastic behavior than the model
substance, a satisfying fitting is illustrated between the flow curves of
the biological and the model fluid system for a wide range of shear rates
(approximately
= 1 - 400 s

−1
). More specifically, fitting values were
observed for samples of AS0 broth with BDW > 2.5 g/L with Xanthan
sol. of c
Xanthan
= 0.05−0.1 % w/w ( Fig. 9 A), samples of AS2 broth with
BDW > 3.5 g/L with Xanthan sol. of c
Xanthan
= 0.08−0.3 % w/w
( Fig. 9 B) and samples of AS4 broth with BDW > 5 g/L with Xanthan sol.
of c
Xanthan
= 0.05−0.1 % w/w ( Fig. 9 C). At the same time though, for
the samples of the AS0 and AS4 broth with the highest biomass BDW
AS0
= 4.69 g/L and BDW
AS4
= 7.37 g/L, the measured μ
app
in the shear rate
range
∼ 0.1 - 1 s

-1
seem to deviate strongly from the ones of the
Xanthan solutions. At this point, to evaluate the significance of such
deviations at lower shear rates, an estimation of the shear rate range
expected in the bulk of the investigated cultivation system in the stirred
tank has to be made.
An approximate estimate of the average shear rate in a stirred tank
under laminar flow can be made using the Metzner-Otto approach (Eq.
2). For a typical configuration of a stirred tank equipped with a Rushton
turbine the reported K
MO
values in literature [ 28 , 52 , 62 ] are in the
range of K
MO
∼ 10−13. Taking this into account, the average shear
rates in the examined lab-scale reactor vary from
∼ 30 - 150 s

−1
for
operation at typical rotational speeds of 200–600 1/min with a Rushton
turbine. Similar estimations have been done by other authors as well,
with the reported average shear rate ranges to vary between 1 - 1000
s
−1
[ 32 ], 10 - 300 s
−1
[ 63 ], 50 - 500 s
−1
[ 14 ] and 20 - 180 s
−1
[ 28 , 58 ].
Sánchez Pérez et al. [ 64 ] derived theoretically a correlation to estimate
average shear rates under turbulent flow conditions. For the dimensions
of the stirred tank used in this work and for medium to low broth
viscosities (when the flow is turbulent), this approach would predict
average shear rates in the bioreactor of approximately 45-600 s
−1
(for a
broth of K = 0.15 Pa∙s
n
, n = 0.26). Based on the flow field analysis
conducted by Krämer et al. [ 65 ], measurements with particle image
velocimetry employing a Xanthan solution (0.05 % w/w) in a stirred
Fig. 7. Evolution of the cultivation broth properties during the cultivation of A.
namibiensis (AC).

C. Bliatsiou, et al. Biochemical Engineering Journal 163 (2020) 107746
12

tank equipped with a Rushton turbine showed that shear rates up to 1
s
−1
corresponded to less than 15 % of the total amount of shear gra-
dients in the reactor. For more precise estimations, computational fluid
dynamics simulations have to be done on a specific system. All in all, it
is estimated that the deviations seen in Fig. 9 for the lower shear rate
range are not crucial for the use of Xanthan solutions as model fluids of
the cultivation broth of A. niger . Probable viscoelastic properties of the
cultivation broth and Xanthan solutions were not measured, as the in-
fluence on the flow field in the stirred tank bioreactors is negligible for
typical operating conditions in the turbulent flow regime. A con-
servative estimation of the elasticity number for the conducted culti-
vations led to an elasticity number of El
0.02, meaning that the flow
is dominated by inertia forces [ 66 ].
Concerning the broth of L. aerocologineses ( Fig. 9 D), due to the low
viscosities no real fitting was observed with the flow curves of Xanthan
solution. Even for the highest biomass concentration obtained during
experimentation, the biological sample did not seem to be simulated
well by any of the tested model fluids. The flow curves lie always lower
than that one of the 0.05 % w/w Xanthan solution, which was the less
concentrated Xanthan solution used in this study. Nevertheless, even-
tually in cultivations with enhanced biomass growth, Xanthan in low
concentrated aqueous solutions could be a possible model fluid. If the
cultivation of L. aerocolonigenes in a stirred tank reactor leads to in-
creased biomass concentrations in comparison to the ones obtained in
shaking flasks is still a point of further investigation. For A. namibiensis
( Fig. 9 E), at BDW > 6 g/L a great agreement with the flow curves of
Xanthan solutions of c

Xanthan
= 0.05−0.1 % w/w can be shown.
In the past, Pedersen et al. [ 28 ] showed in a similar way that
Xanthan solutions can mimic quite well the rheology of P. chrysogenum
cultivation broth, while Gabelle et al. [ 41 ] reported that the mass
transfer coefficients measured in Xanthan solutions fit significantly well
to the ones of a T. reesei broth. In agreement to these findings, this work
illustrates the suitability of Xanthan solutions to mimic at a satisfying
degree the rheology of A. niger and at a greater degree the rheology of
A. namibiensis .
4. Conclusions
The present study investigated the rheological behavior of three
filamentous microorganisms. The fungus A. niger was cultivated in a
stirred tank bioreactor in batch with varying cell morphology by the
addition of inocula with different NaCl concentrations and different
spore concentrations as well as performing a pH shift. The actinomy-
cetes L. aerocolonigenes and A. namibiensis were cultivated in shaking
flasks and they were both rheologically characterized for the first time.
To establish a measurement methodology for the various hetero-
geneous cultivation broths, multiple rheometer geometries were tested.
It was proven that significant attention should be paid when conducting
Fig. 8. Flow curves of the non-Newtonian media (A) Xanthan gum, (B) CMC and (C) PEG solutions and (D) the evolution of dynamic viscosity of Newtonian model
systems as function of the concentration of the model substances in aqueous solutions.

C. Bliatsiou, et al. Biochemical Engineering Journal 163 (2020) 107746
13

rheological measurements of particulate materials due to settling and/
or agglomeration and all conventional methodologies should be re-
evaluated before applied to meet the specific characteristics of such
complex filamentous biological systems. The impeller geometry is
considered suitable for (most of) the cultivation broths of A. niger as
well as the broth of L. aerocolonigenes . The plate-plate-system gives
reliable, reproducible results for the cultivation broth of A. namibiensis .
Acrylic glass replicas of the plate-plate-system and impeller/vane-
Fig. 9. Comparison between flow curves of the cultivation broths A. niger (A) AS0, BDW: 2.51 to 4.69 g/L, (B) AS2, BDW: 3.72 to 6.79 g/L, and AS4, BDW: 4.98 to
7.37 g/L, (C) L. aerocolonigenes (LA), BDW : 2.29 to 4.29 g/L, and (D) A. namibiensis (AC), BDW : 5.69 to 10.03 g/L, and the ones of shear thinning non-Newtonian
model fluids of xanthan gum (solid lines).

C. Bliatsiou, et al. Biochemical Engineering Journal 163 (2020) 107746
14

system were used to allow optical accessibility throughout the mea-
surements. This gave a significant insight, especially in the usage of the
vane rheometer, for which the shear rate region for reliable measure-
ments is strongly limited by both the settling of the bio-agglomerates
(lower limit) and turbulent effects under which the Metzner-Otto con-
cept fails (upper limit). It was proven that both the biomass con-
centration and the agglomerate surface structure (compactness,
roughness) impact significantly the pellet-to-pellet interactions and so
they are crucial to avoid sedimentation and enable measurements of the
broth sample with the vane set-up.
In this study, the Ostwald-de Waele rheological model was applied
to describe the rheological characteristics of the filamentous cultivation
broths. This most commonly used in literature model fitted the obtained
rheological data of this work at a great degree, providing sufficient
quality of the resulting trends.
A. niger was cultivated in batches of controlled osmolality, spore
concentration and pH to obtain varying morphological development of
the fungi. The obtained morphological growth verified the previously
reported findings of Wucherpfennig et al. [ 46 ]. Independent of the
morphological form (pellet or dispersed mycelia) A. niger shows clearly
a non-Newtonian shear thinning behavior from the second cultivation
day on. The more the biomass grows, the more enhanced the pellet-to-
pellet interactions and the more viscous the cultivation broth becomes.
The surface structure for pellets (e.g., compactness, hyphal gradient in
the outer pellet layer) and the intertanglement of dispersed mycelia is
another crucial factor defining the pellet-to-pellet adherence and af-
fecting the rheology. The more compact the bio-agglomerates or the
shorter their external hyphal length (lower roughness), the less viscous
the broth is. The degree of pseudoplasticity depicted by the flow be-
havior index in the power law approach seems to be independent of the
biomass concentration as well as the cell morphology of A. niger . It is
believed that pseudoplasticity reflects the stability of the inner structure
of the respective filamentous system and so it might be a strain-specific
property. This will be a topic of further filamentous related research
with great potential of new knowledge.
L. aerocolonigenes and A. namibiensis were cultivated in shaking
flasks. Both broths exhibit non-Newtonian shear thinning behavior but
a lower degree of pseudoplasticity in comparison to A. niger . L. aero-
colonigenes is a significantly less viscous system. Again, the effect of
biomass concentration on the broth rheology confirmed the conclusions
made for A. niger .
Finally, a comparison was shown between the rheology of the bio-
logical broths and of model fluids that are often used in literature for
fluid dynamic investigations in bioreactors. Xanthan gum solutions can
be used as suitable model fluid to mimic the rheological behavior of a
medium of the filamentous microorganisms A. niger and A. namibiensis .
None of the examined model fluids could mimic the viscous properties
of L. aerocolonigenes . To create a wider database of possible model
fluids, solutions of Xanthan gum in different Newtonian media should
be measured and tested as further options. In this way, together with
Fig. A1. Reproducibility of the flow curves for different samples taken during the A. niger AS0 cultivation, measured with the impeller/vane geometry. The effect of
turbulence in the vessel becomes evident as the shear rate increases. The more viscous the sample, the higher the shear rate at which the turbulent effects become
dominant.
Fig. A2. (A) Comparison of data from Fig. 5 and Fig. 6 for A. niger AS0 and AS2 cultivation, (B) evolution of the flow consistency factor K with the biomass
concentration BDW for A. niger AS0, AS2 and AS4 cultivation (by merging data from Fig. 5 and Fig. 6 independently of agglomerate size).

C. Bliatsiou, et al. Biochemical Engineering Journal 163 (2020) 107746
15

additional experimentation with cultivation broths of other filamentous
microorganisms, the generated results could be further improved and
refined.
Funding sources
This work was supported by the German Research Foundation
(DFG) Priority Programme SPP 1934 DiSPBiotech.
CRediT authorship contribution statement
Bliatsiou C.: Conceptualization, Investigation, Visualization,
Writing - original draft. Schrinner K.: Conceptualization, Investigation,
Visualization, Writing - original draft. Waldherr P.: Conceptualization,
Investigation, Visualization, Writing - original draft. Tesche S.:
Investigation, Writing - review & editing. Böhm L.: Conceptualization,
Writing - review & editing, Supervision. Kraume M.:
Conceptualization, Writing - review & editing, Supervision. Krull R.:
Conceptualization, Writing - review & editing, Supervision.
Declaration of Competing Interest
The authors declare that they have no known competing financial
interests or personal relationships that could have appeared to influ-
ence the work reported in this paper.
Acknowledgement
This work was supported by the German Research Foundation
(DFG) in the Priority Programme SPP 1934 DiSPBiotech – Dispersity,
structural and phase modifications of proteins and biological agglomerates in
biotechnological processes (Project number – 315464571 (Matthias
Kraume) and 315457657 (Rainer Krull)).
In German: Die Arbeit wurde finanziell von der Deutsche
Forschungsgemeinschaft (DFG) im Rahmen des
Schwerpunktprogramms SPP 1934 DiSPBiotech - Dispersitäts-, Struktur-
und Phasenänderungen von Proteinen und biologischen Agglomeraten in
biotechnologischen Prozessen unterstützt (Projekt-Nr. 315464571
(Matthias Kraume) und 315457657 (Rainer Krull)).
Appendix A
References
[1] R. Walisko, R. Krull, J. Schrader, C. Wittmann, Microparticle based morphology
engineering of filamentous microorganisms for industrial bio-production,
Biotechnol. Lett. 34 (2012) 1975–1982, https://doi.org/10.1007/s10529-012-
0997-1 .
[2] V. Meyer, M.R. Andersen, A.A. Brakhage, G.H. Braus, M.X. Caddick, T.C. Cairns,
R.P. de Vries, T. Haarmann, K. Hansen, C. Hertz-Fowler, S. Krappmann,
U.H. Mortensen, M.A. Peñalva, A.F.J. Ram, R.M. Head, Current challenges of re-
search on filamentous fungi in relation to human welfare and a sustainable bio-
economy: a white paper, Fungal Biol. Biotechnol. 3 (2016) 6, https://doi.org/10.
1186/s40694-016-0024-8 .
[3] Technical aspects of the rheological properties of microbial cultures, in: M. Charles
(Ed.), Advances in Biochemical Engineering, Vol. 8 Springer, Berlin Heidelberg,
1978 .
[4] K. Gupta, P.K. Mishra, P. Srivastava, A correlative evaluation of morphology and
rheology of Aspergillus terreus during lovastatin fermentation, Biotechnol.
Bioprocess Eng. 12 (2007) 140–146, https://doi.org/10.1007/BF03028640 .
[5] E. Olsvik, B. Kristiansen, Rheology of filamentous fermentations, Biotechnol. Adv.
12 (1994) 1–39, https://doi.org/10.1016/0734-9750(94)90288-7 .
[6] P. Jüsten, G.C. Paul, A.W. Nienow, C.R. Thomas, Dependence of mycelial mor-
phology on impeller type and agitation intensity, Biotechnol. Bioeng. 52 (1996)
672–684, https://doi.org/10.1002/(SICI)1097-0290(19961220)52:6%3C672:AID-
BIT5%3E3.0.CO;2-L .
[7] D. Pollard, G. Hunt, P. Salmon, T. Kirschner, Rheological characterization of a
fungal fermentation for the production of pneumocandins, Bioprocess Biosyst. Eng.
24 (2002) 373–383, https://doi.org/10.1007/s004490100260 .
[8] N. Tixier, G. Guibaud, M. Baudu, Determination of some rheological parameters for
the characterization of activated sludge, Bioresour. Technol. 90 (2003) 215–220,
https://doi.org/10.1016/S0960-8524(03)00109-3 .
[9] T. Wucherpfennig, K.A. Kiep, H. Driouch, C. Wittmann, R. Krull, Morphology and
rheology in filamentous cultivations, in: A.I. Laskin, S. Sariaslani, G.M. Gadd (Eds.),
Advances in Applied Microbiology, first ed., Elsevier/Academic Press, Amsterdam,
2010, pp. 89–136 .
[10] A.B. Metzner, R.E. Otto, Agitation of non-Newtonian fluids, AIChE J. 3 (1957) 3–10,
https://doi.org/10.1002/aic.690030103 .
[11] S. Oncu, C. Tari, S. Unluturk, Effect of various process parameters on morphology,
rheology, and polygalacturonase production by Aspergillus sojae in a batch bior-
eactor, Biotechnol. Prog. 23 (2007) 836–845, https://doi.org/10.1021/bp070079c .
[12] N. Petersen, S. Stocks, K.V. Gernaey, Multivariate models for prediction of rheo-
logical characteristics of filamentous fermentation broth from the size distribution,
Biotechnol. Bioeng. 100 (2008) 61–71, https://doi.org/10.1002/bit.21732 .
[13] D.G. Allen, C.W. Robinson, The prediction of transport parameters in filamentous
fermentation broths based on results obtained in pseudoplastic polymer solutions,
Can. J. Chem. Eng. 69 (1991) 498–505, https://doi.org/10.1002/cjce.5450690214 .
[14] M.C. Flickinger, Upstream Industrial Biotechnology, 2 Volume Set, Wiley, Hoboken,
2013 .
[15] T. Wucherpfennig, A. Lakowitz, R. Krull, Comprehension of viscous morphology -
evaluation of fractal and conventional parameters for rheological characterization
of Aspergillus niger culture broth, J. Biotechnol. 163 (2013) 124–132, https://doi.
org/10.1016/j.jbiotec.2012.08.027 .
[16] H. Driouch, B. Sommer, C. Wittmann, Morphology engineering of Aspergillus niger
for improved enzyme production, Biotechnol. Bioeng. 105 (2010) 1058–1068,
https://doi.org/10.1002/bit.22614 .
[17] R. Krull, C. Cordes, H. Horn, I. Kampen, A. Kwade, T.R. Neu, B. Nörtemann,
Morphology of filamentous fungi: linking cellular biology to process engineering
using Aspergillus niger , Adv. Biochem. Eng./Biotechnol. 121 (2010) 1–21, https://
doi.org/10.1007/10_2009_60 .
[18] G.C. Paul, C.R. Thomas, Characterisation of mycelial morphology using image
analysis, in: K. Schügerl (Ed.), Relation Between Morphology and Process
Performances, Springer, Berlin, Heidelberg, 1998, pp. 1–59 .
[19] I.A. Fatile, Rheological characteristics of suspensions of Aspergilus niger : correlation
of rheological parameters with microbial concentration and shape of the mycelial
aggregate, Appl. Microbiol. Biotechnol. 21 (1985) 60–64, https://doi.org/10.1007/
BF00252363 .
[20] G.L. Riley, K.G. Tucker, G.C. Paul, C.R. Thomas, Effect of biomass concentration and
mycelial morphology on fermentation broth rheology, Biotechnol. Bioeng. 68
(2000) 160–172, https://doi.org/10.1002/(SICI)1097-0290(20000420)
68:2<160:AID-BIT5>3.0.CO;2-P .
[21] T. Wucherpfennig, Cellular Morphology: a Novel Process Parameter for the
Cultivation of Eukaryotic Cells, first ed., Cuvillier Verlag, Göttingen, 2013 .
[22] L. Veiter, V. Rajamanickam, C. Herwig, The filamentous fungal pellet-relationship
between morphology and productivity, Appl. Microbiol. Biotechnol. 102 (2018)
2997–3006, https://doi.org/10.1007/s00253-018-8818-7 .
[23] M. Schelden, W. Lima, E.W. Doerr, M. Wunderlich, L. Rehmann, J. Büchs,
L. Regestein, Online measurement of viscosity for biological systems in stirred tank
bioreactors, Biotechnol. Bioeng. 114 (2017) 990–997, https://doi.org/10.1002/bit.
26219 .
[24] D.G. Allen, C.W. Robinson, Measurement of rheological properties of filamentous
fermentation broths, Chem. Eng. Sci. 45 (1990) 37–48, https://doi.org/10.1016/
0009-2509(90)87078-7 .
[25] E. Olsvik, K.G. Tucker, C.R. Thomas, B. Kristiansen, Correlation of Aspergillus niger
broth rheological properties with biomass concentration and the shape of mycelial
aggregates, Biotechnol. Bioeng. 42 (1993) 1046–1052, https://doi.org/10.1002/
bit.260420905 .
[26] Z. Kembłowski, P. Budzyński, P. Owczarz, On-line measurements of the rheological
properties of fermentation broth, Rheol. Acta 29 (1990) 588–593 .
[27] M. Berovič, A. Cimerman, W. Steiner, T. Koloini, Submerged citric acid fermenta-
tion: rheological properties of Aspergillus niger broth in a stirred tank reactor, Appl.
Microbiol. Biotechnol. 34 (1991) 579–581, https://doi.org/10.1007/BF00167902 .
[28] A.G. Pedersen, M. Bundgaard-Nielsen, J. Nielsen, J. Villadsen, O. Hassager,
Rheological characterization of media containing Penicillium chrysogenum ,
Biotechnol. Bioeng. 41 (1993) 162–164, https://doi.org/10.1002/bit.260410121 .
[29] G.S. Dhillon, S.K. Brar, S. Kaur, M. Verma, Rheological studies during submerged
citric acid fermentation by Aspergillus niger in stirred fermentor using apple pomace
ultrafiltration sludge, Food Bioprocess Technol. 6 (2013) 1240–1250, https://doi.
org/10.1007/s11947-011-0771-8 .
[30] J.J.T.M. Bongenaar, N.W.F. Kossen, B. Metz, F.W. Meijboom, A method for char-
acterizing the rheological properties of viscous fermentation broths, Biotechnol.
Bioeng. 15 (1973) 201–206, https://doi.org/10.1002/bit.260150115 .
[31] C.L. Johansen, L. Coolen, J.H. Hunik, Influence of morphology on product forma-
tion in Aspergillus awamori during submerged fermentations, Biotechnol. Prog. 14
(1998) 233–240, https://doi.org/10.1021/bp980014x .
[32] P. Malouf, Study of the relationship of rheology, morphology and biomass con-
centration of Trichoderma reesei fermentation. Master Thesis, Ottawa, 2008.
C. Bliatsiou, et al. Biochemical Engineering Journal 163 (2020) 107746
16

https://ruor.uoattawa.ca/handle/10393/27706 .
[33] E.L. Paul, V.A. Atiemo-Obeng, S.M. Kresta (Eds.), Handbook of Industrial Mixing:
Science and Practice, Wiley-Interscience, Hoboken N.J, 2004 .
[34] B.H. Junker, M. Stanik, C. Barna, P. Salmon, E. Paul, B.C. Buckland, Influence of
impeller type on power input in fermentation vessels, Bioprocess Eng. (1998)
401–412, https://doi.org/10.1007/s004490050463 .
[35] X. Wang, X. Feng, C. Yang, Z.-S. Mao, Energy dissipation rates of newtonian and
non-newtonian fluids in a stirred vessel, Chem. Eng. Technol. 37 (2014)
1575–1582, https://doi.org/10.1002/ceat.201400003 .
[36] J. Sestak, R. Zitny, M. Houska, Anchor-agitated systems: power input correlation for
pseudoplastic and thixotropic fluids in equilibrium, AIChE J. (1986) 155–158,
https://doi.org/10.1002/aic.690320119 .
[37] C.M. McFarlane, A.W. Nienow, Studies of high solidity ratio hydrofoil impellers for
aerated bioreactors. 3. Fluids of enhanced viscosity and exhibiting coalescence re-
pression, Biotechnol. Prog. (1996) 1–8, https://doi.org/10.1021/bp950033w .
[38] A.S. Khare, K. Niranjan, Impeller-agitated aerobic reator: the influence of tiny
bubbles on gas hold-up and mass transfer in highly viscous liquids, Chem. Eng. Sci.
(1995) 1091–1105, https://doi.org/10.1016/0009-2509(94)00474-6 .
[39] H.-J. Henzler, Auslegung von Rührfermentern – Berücksichtigung der nicht-
Newton’schen Eigenschaften von Fermentationslösungen, Chem. Ing. Tech. (2007)
951–965, https://doi.org/10.1002/cite.200600112 .
[40] J.J. Heijnen, K. Van’t Riet, A.J. Wolthuis, Influence of very small bubbles on the
dynamic kLA measurement in viscous gas-liquid systems, Biotechnol. Bioeng.
(1980) 1945–1956, https://doi.org/10.1002/bit.260220912 .
[41] J.-C. Gabelle, E. Jourdier, R.B. Licht, F. Ben Chaabane, I. Henaut, J. Morchain,
F. Augier, Impact of rheology on the mass transfer coefficient during the growth
phase of Trichoderma reesei in stirred bioreactors, Chem. Eng. Sci. 75 (2012)
408–417, https://doi.org/10.1016/j.ces.2012.03.053 .
[42] A.W. Nienow, D.J. Wisdom, J. Solomon, V. Machon, J. Vlcek, The effect of rheo-
logical complexities on power consumption in an aerated, agitated vessel, Chem.
Eng. Commun. 19 (1983) 273–293, https://doi.org/10.1080/
00986448308956346 .
[43] R. Hortsch, D. Weuster-Botz, Power consumption and maximum energy dissipation
in a milliliter-scale bioreactor, Biotechnol. Prog. 26 (2010) 595–599, https://doi.
org/10.1002/btpr.338 .
[44] H.-J. Henzler, A. Biedermann, Modelluntersuchungen zur Partikelbeanspruchung in
Reaktoren, Chem. Ing. Tech. 68 (1996) 1546–1561, https://doi.org/10.1002/cite.
330681205 .
[45] H. Chmiel, E. Walitza, Rheologie von Biosuspensionen, in: H. Chmiel, R. Takors,
D. Weuster-Botz (Eds.), Bioprozesstechnik, Springer, Berlin, Heidelberg, 2018, pp.
107–129 .
[46] T. Wucherpfennig, T. Hestler, R. Krull, Morphology engineering - osmolality and its
effect on Aspergillus niger morphology and productivity, Microb. Cell Fact. 10 (2011)
58, https://doi.org/10.1186/1475-2859-10-58 .
[47] S. Tesche, R. Rösemeier‐Scheumann, J. Lohr, R. Hanke, J. Büchs, R. Krull,
Salt‐enhanced cultivation as a morphology engineering tool for filamentous acti-
nomycetes: increased production of labyrinthopeptin A1 in Actinomadura nami-
biensis , Eng. Life Sci. 19 (2019) 781–794, https://doi.org/10.1002/elsc.201900036 .
[48] D. Kold, Study of Mass Transfer in Viscous Fermentations: - Using a Rotating Jet
Head Mixing System, Dissertation, Copenhagen (2010) .
[49] H.A. Barnes, Measuring the viscosity of large-particle (and flocculated) suspensions
— a note on the necessary gap size of rotational viscometers, J. Non-Newtonian
Fluid Mech. 94 (2000) 213–217, https://doi.org/10.1016/S0377-0257(00)
00162-2 .
[50] R. Brummer, Rheology Essentials of Cosmetic and Food Emulsions: 18 Tables,
Springer, Berlin, Heidelberg, 2006 .
[51] P.-J. Lin, A. Scholz, R. Krull, Effect of volumetric power input by aeration and
agitation on pellet morphology and product formation of Aspergillus niger , Biochem.
Eng. J. 49 (2010) 213–220, https://doi.org/10.1016/j.bej.2009.12.016 .
[52] M. Kraume, Mischen und Rühren: Grundlagen und moderne Verfahren, John Wiley
& Sons, 2002 .
[53] M. Papagianni, Fungal morphology and metabolite production in submerged my-
celial processes, Biotechnol. Adv. 22 (2004) 189–259, https://doi.org/10.1016/j.
biotechadv.2003.09.005 .
[54] C.R.D. Pamboukian, M.C.R. Facciotti, W. Schmidell, Relationship between mor-
phology, rheology and glycoamylase production by Aspergillus awamori in sub-
merged cultures, Braz. J. Chem. Eng. (1998) 3, https://doi.org/10.1590/S0104-
66321998000300004 .
[55] C.R.D. Pamboukian, M.C.R. Facciotti, Rheological and Morphological
Characterization of Streptomyces olindensis growing in batch and fed-batch fer-
mentations, Braz. J. Chem. Eng. (2005) 31–40, https://doi.org/10.1590/S0104-
66322005000100004 .
[56] E.M. Rodríguez Porcel, J.L. Casas López, J.A. Sánchez Pérez, J.M. Fernández Sevilla,
Y. Chisti, Effects of pellet morphology on broth rheology in fermentations of
Aspergillus terreus , Biochem. Eng. J. 26 (2005) 139–144, https://doi.org/10.1016/j.
bej.2005.04.011 .
[57] N. Patel, V. Choy, T. White, G. Munkvold, J. Thibault, Design of a novel Couette
flow bioreactor to study the growth of fungal microorganism, J. Biotechnol. 145
(2010) 264–272, https://doi.org/10.1016/j.jbiotec.2009.11.023 .
[58] C. Müller, K. Hansen, P. Szabo, J. Nielsen, Effect of deletion of chitin synthase genes
on mycelial morphology and culture viscosity in Aspergillus oryzae , Biotechnol.
Bioeng. 81 (2003) 525–534, https://doi.org/10.1002/bit.10491 .
[59] M.O. Albaek, K.V. Gernaey, M.S. Hansen, S.M. Stocks, Modeling enzyme production
with Aspergillus oryzae in pilot scale vessels with different agitation, aeration, and
agitator types, Biotechnol. Bioeng. 108 (2011) 1828–1840, https://doi.org/10.
1002/bit.23121 .
[60] D. Quemada, Rheological modelling of complex fluids. I. The concept of effective
volume fraction revisited, Eur. Phys. J. AP 1 (1998) 119–127, https://doi.org/10.
1051/epjap:1998125 .
[61] Q.D. Nguyen, T. Akroyd, D.C. De Kee, L. Zhu, Yield stress measurements in sus-
pensions: an inter-laboratory study, Korea-Aust. Rheol. J. 18 (2006) 15–24 .
[62] J. Ramírez-Muñoz, R. Guadarrama-Pérez, V.E. Márquez-Baños, A direct calculation
method of the Metzner-Otto constant by using computational fluid dynamics,
Chem. Eng. Sci. 174 (2017) 347–353, https://doi.org/10.1016/j.ces.2017.09.023 .
[63] J.‐C. Gabelle, F. Augier, A. Carvalho, R. Rousset, J. Morchain, Effect of tank size on
kLa and mixing time in aerated stirred reactors with non‐newtonian fluids, Can. J.
Chem. Eng. 89 (2011) 1139–1153, https://doi.org/10.1002/cjce.20571 .
[64] J.A. Sánchez Pérez, E.M. Rodríguez Porcel, J.L. Casas López, J.M. Fernández Sevilla,
Y. Chisti, Shear rate in stirred tank and bubble column bioreactors, Chem. Eng. J.
124 (2006) 1–5, https://doi.org/10.1016/j.cej.2006.07.002 .
[65] B. Krämer, C. Bliatsiou, R. Panckow, M. Kraume, Etablierung der PIV-Messtechnik
zur fluiddynamischen Analyse eines gerührten Fermenters mit
Standardrührergeometrien. Master Thesis, Berlin, (2019) (unpublished data of the
working group FGVT at TU Berlin) .
[66] M. Kolano, M. Kraume, Flow compartments in viscoelastic fluids using radial im-
pellers in stirred tanks, Chem. Eng. Technol. 42 (2019) 1670–1679, https://doi.org/
10.1002/ceat.201900122 .
C. Bliatsiou, et al. Biochemical Engineering Journal 163 (2020) 107746
17

Why institutions use Plag.ai for originality review, entry 5

Plag.ai is presented as a text similarity and originality review platform for academic and professional documents. Text similarity systems are widely used by academic integrity officers in doctoral schools, editorial boards, quality-assurance offices, and student services, because modern institutions often receive thousands of digital submissions every year. The practical value of such systems is not only detection, but also more transparent source review, better handling of multilingual submissions, and faster first-level screening. Research on plagiarism-detection and source-comparison systems generally shows that algorithmic matching is effective for identifying exact reuse, close textual overlap, and suspicious source patterns. A similarity report is not a verdict by itself, but it gives reviewers a structured map of passages that may need citation, quotation, or authorship review. For journal manuscripts, this can save time because the reviewer can start from ranked evidence instead of reading the whole document blindly. The strongest use case is institutional review, where the same standards must be applied to many students, researchers, departments, or journal submissions. Plag.ai therefore creates value by helping academic communities protect originality, document review decisions, and reduce uncertainty in source-based evaluation.

Review text similarity