Mapping Nanocellulose- and Alginate-Based Photosynthetic Cell
Factory Scaffolds: Interlinking Porosity, Wet Strength, and Gas
Exchange
Tuukka Levä,⊥ Ville Rissanen,*,⊥ Lauri Nikkanen,⊥ Vilja Siitonen, Maria Heilala, Josphat Phiri,
Thaddeus C. Maloney, Sergey Kosourov, Yagut Allahverdiyeva, Mikko Mäkelä, and Tekla Tammelin*
Cite This: Biomacromolecules 2023, 24, 3484−3497 Read Online
ACCESS Metrics & More Article Recommendations *sı Supporting Information
ABSTRACT: To develop efficient solid-state photosynthetic cell factories for
sustainable chemical production, we present an interdisciplinary experimental toolbox
to investigate and interlink the structure, operative stability, and gas transfer properties
of alginate- and nanocellulose-based hydrogel matrices with entrapped wild-type
Synechocystis PCC 6803 cyanobacteria. We created a rheological map based on the
mechanical performance of the hydrogel matrices. The results highlighted the
importance of Ca2+-cross-linking and showed that nanocellulose matrices possess
higher yield properties, and alginate matrices possess higher rest properties. We
observed higher porosity for nanocellulose-based matrices in a water-swollen state via
calorimetric thermoporosimetry and scanning electron microscopy imaging. Finally, by
pioneering a gas flux analysis via membrane-inlet mass spectrometry for entrapped cells,
we observed that the porosity and rigidity of the matrices are connected to their gas
exchange rates over time. Overall, these findings link the dynamic properties of the life-
sustaining matrix to the performance of the immobilized cells in tailored solid-state
photosynthetic cell factories.
1. INTRODUCTION
Whole-cell immobilization of different cell types ranging from
heterotrophic microbes and animal cells to photosynthetic
organisms has been extensively studied for a variety of
applications including tissue engineering,1,2 3D cell culture
and analysis matrices,3−5 wastewater purification,6−8 and
biotransformation or production of chemicals.9−12 In these
biohybrid platforms, the properties of both the cells and the
matrix are pivotal factors affecting the operational performance
of the system as a whole. Cell-laden hydrogels are most often
used in the biomedical field and have generally been
characterized using imaging techniques that can preserve the
sample ultrastructure (e.g., scanning electron microscopy
(SEM), confocal laser scanning microscopy, and atomic force
microscopy (AFM)),13 mechanical assessments (compressive
and tensile testing, rheology, swelling behavior, porosity, and
gel degradation),14,15 and by evaluating the biological
compatibility of the system (cell viability, growth, and
morphology).10,11,15 The selection of methods is heavily
dependent on the end application, which defines the
operational framework and boundary conditions for the
applied materials.
In this article, specifically, we investigate the development of
solid-state photosynthetic cell factories (SSPCFs). SSPCFs are
versatile platforms for efficient and sustainable production of
targeted chemicals, where photosynthetic cells are immobilized
via entrapment within a thin and transparent hydrogel matrix.
From the biological perspective, various photosynthetic
production hosts have been extensively studied, and their
robustness and productivity have been modified and enhanced
with cell and metabolic engineering.16−20 Both the prokaryotic
cyanobacteria and the diverse group of eukaryotic microalgae
include species that have been identified as potential
photosynthetic cell factory hosts, with a wide product range
from biofuels to specialty chemicals.21−24 Furthermore, many
investigations on suitable immobilization mechanisms and
matrix components have been conducted.9,25−27 From the
matrix perspective, hydrogels based on cross-linked alginate
have been identified as well-performing materials with
sustainable natural sources and high biocompatibility.10,28−31
Recently, hydrogels fabricated from nanocellulosic materials,
such as bacterial cellulose,32 carboxymethylated cellulose
nanofibrils,33 and TEMPO-oxidized cellulose nanofibers
(TCNF),11,34 have been demonstrated as promising alter-
Received: March 13, 2023
Revised: June 20, 2023
Published: June 29, 2023
Articlepubs.acs.org/Biomac
© 2023 The Authors. Published by
American Chemical Society
3484
https://doi.org/10.1021/acs.biomac.3c00261
Biomacromolecules 2023, 24, 3484−3497
D
ow
nl
oa
de
d 
vi
a 
U
N
IV
 O
F 
TU
RK
U
 o
n 
A
ug
us
t 2
1,
 2
02
3 
at
 0
4:
31
:3
9 
(U
TC
).
Se
e 
ht
tp
s:/
/p
ub
s.a
cs
.o
rg
/sh
ar
in
gg
ui
de
lin
es
 fo
r o
pt
io
ns
 o
n 
ho
w
 to
 le
gi
tim
at
el
y 
sh
ar
e 
pu
bl
ish
ed
 a
rti
cl
es
.
natives for conventional matrix structures due to their
availability, mechanical properties, and modifiability to create
tailored structures. We have previously performed a compar-
ison between cross-linked TCNF and alginate hydrogels with
regard to their mechanical performance in submerged
production conditions.11 However, there is still a distinct
need for interdisciplinary efforts combining materials science,
cell biology, and bioprocess engineering to assess how the
matrix properties are linked to the overall performance of the
SSCPF.
Here, we bring together a strong multidisciplinary expertise
combining materials engineering, chemometrics, biotechnol-
ogy, and photosynthesis research to form a thorough
investigation of these matrix materials with many experimental
methods that are novel to the SSPCF concept and its
development. We present a systematic investigation of the
structure, mass transfer, and operative stability of TCNF- and
alginate-based SSPCF matrices and mixtures thereof (Figure
1). These materials have been investigated in detail in the
context of many other application areas.5,35−39 However, the
characterization techniques are often not directly applicable to
SSPCF matrix scaffolds, where biological compatibility with
the entrapped production hosts necessitates a very specific
interplay between different material properties such as
hydrogel strength, wet-state porosity, and gas transfer
efficiency. To overcome this lack of established methodology,
we identified an experimental toolbox that could characterize
the relevant properties of this unique system. This toolbox was
used for the mapping of the matrix materials for constructing
versatile SSPCFs with tailored properties. Moreover, as cross-
linking of alginate and TCNF is required to obtain self-
standing matrices, we deepened the investigation of the effect
of calcium-40 and PVA-based41 cross-linking approaches on the
matrices.
We employed small deformation oscillatory rheology to
investigate the wet strength of the matrices under shearing. We
then determined the systematic variations and groupings in
matrix properties with principal component analysis (PCA)42
and unsupervised hierarchical clustering.43 As a result, we were
able to highlight the important rheological features governing
the matrix compositions and cross-linking types. To observe
the porosity of the matrices in the wet state, we used a
technique called thermoporosimetry (TPM), which is based
on calorimetric scanning and has been employed to calculate
the pore volumes and size distributions of mesoporous
materials including nanocellulose networks.44−46 The results
from thermoporosimetry were complemented by imaging the
matrix surfaces with scanning electron microscopy (SEM).
Finally, we entrapped Synechocystis sp. PCC 6803 wild-type
cyanobacterial cells in the matrices and assessed the relation-
ship between the matrix porosity and the biological perform-
ance of the immobilized cells by following the gas transfer of
the matrices. The kinetics and net gas exchange of CO2 and O2
during controlled periods of light and dark were investigated
with membrane inlet mass spectrometry (MIMS), which has
not been used before in the research aiming at developing
SSPCFs.
We found that despite being confined to operate near the
limits of our chosen techniques, we were able to observe
distinctions in the initial wet strength, porosity, and gas
transfer properties between TCNF- and alginate-based
matrices. The mixed formulations exhibited properties mostly
from the dominant component. Additionally, we identified the
use of calcium ions as an effective cross-linking mechanism for
both TCNF and alginate. Notably, the different interactions
between the matrix components and the ions lead to diverse
material behavior that can attribute to dynamic changes in
matrix properties. Overall, the interdisciplinary methodological
toolbox presented in this work provides a means to better
understand the complex and interdependent relationship
between matrix properties and SSPCF performance. This
knowledge is required to develop tailorable cell immobilization
matrices that can be fitted to the needs of specific production
hosts and operating conditions.
Figure 1. Concept of this study featuring the key dimensions and parameters for the development of solid-state photosynthetic cell factories
(SSPCFs).
Biomacromolecules pubs.acs.org/Biomac Article
https://doi.org/10.1021/acs.biomac.3c00261
Biomacromolecules 2023, 24, 3484−3497
3485
2. MATERIALS AND METHODS
2.1. Cyanobacterial Cells and Growth Conditions. Wild-type
Synechocystis sp. PCC 6803 cyanobacterial cells (herein denoted as
Synechocystis) were cultivated in the BG-11 growth medium47 that was
buffered with 5 mM HEPES-NaOH (pH 7.5). The cell cultures were
inoculated to 250 mL conical flasks in a total culture volume of 50 mL
and placed on a rotary shaker (110 rpm). The suspension cultures
were incubated at 23 °C under ambient CO2 levels and 16 h of
photoperiod illuminated by fluorescent lamps (Philips Master TL5
HO 39W/865) supplying the cultivated photosynthetic cells with
approximately 50 μmol photons m−2 s−1 photosynthetically active
radiation (PAR). The cultures were periodically renewed in an open
laminar flow cabinet (KOJAIR) after growing them for 7−14 days.
2.2. TEMPO-Oxidized Cellulose Nanofibers (TCNF). TEMPO-
oxidized cellulose nanofibers were manufactured from never-dried
bleached softwood pulp that was obtained from a coniferous wood
mixture consisting of spruce and pine. Pulping was carried out in a
Finnish pulp mill, and the TEMPO-catalyzed oxidation of the pulp
was conducted with alkaline hypochlorite as the primary oxidant
according to the protocol reported by Saito et al.48 2,2,6,6-
Tetramethylpiperidine-1-oxyl (TEMPO) (Sigma-Aldrich) and 10%
sodium hypochlorite (5 mmol/g pulp fiber, Sigma-Aldrich) were used
in the TEMPO-catalyzed oxidation of the pulp. After pulping and
oxidation, an anionic charge of 1.45−1.52 mmol/g was obtained for
the oxidized pulp via a standard conductometric titration method
(SCAN 65:02).49 The oxidized pulp was subsequently washed and
passed through a microfluidizer carrying two Z-type chambers with
respective diameters of 400 and 100 μm (Microfluidics Int., USA)
twice at 1850 bar to fibrillate the oxidized pulp into TCNF. TCNF of
∼1 wt % with viscous gel-like characteristics and transparent optical
properties was obtained. The chemical composition, morphology, and
visual appearance of similarly prepared TCNF-grade used in this
study were described in earlier publications.11,34
2.3. Polymers. Two polymeric components, alginate (ALG) and
polyvinyl alcohol (PVA), were used in addition to TCNF in the
preparation of hydrogel cell immobilization matrices. Alginic acid
sodium salt from brown algae (#71238, Sigma-Aldrich) was used in
this study, with a β-D-mannuronic acid (M) content of 25−35%, α-L-
guluronic acid (G) content of 65−70% (M/G ratio 0.43), and 100−
200 kg/mol Mw, as approximated by the manufacturer. Alginate was
dissolved in Milli-Q water overnight with magnetic stirring to prepare
stock solution of approximately 2 wt %. PVA (Mowiol 56-98, Mw:
195 kg/mol, DP 4300, Sigma-Aldrich), and stock solution of around 5
wt % was prepared by dissolving solid PVA crystals to Milli-Q water
for approximately 1 h until fully dissolved. The water solubility of
PVA was enhanced by stirring and heating the solution to >90 °C in a
water bath during preparation.
2.4. Other Materials. Ultrapure Milli-Q water (18.2 MΩ cm)
used throughout this study in the preparation of solutions and
samples was obtained with a Milli-Q purification unit (QPAK 1,
Millipore). Calcium chloride (50 mM) (CaCl2) solution was prepared
by dissolving CaCl2 of analytical grade (99%, #C7902, Sigma-Aldrich)
to Milli-Q water. A commercial Teflon (PTFE) film (Etra, Finland)
was used as a scaffold during the preparation of all hydrogel films.
2.5. Preparation of Cell Immobilization Matrices. Seven types
of different hydrogel matrices with varying cross-linkers were
prepared. All matrices were prepared to an approximately final 1 wt
% solid content and 1 ± 0.5 mm thickness prior to immersion in
either Milli-Q water (matrices with no cells) or BG-11 growth
medium (matrices with cells). Samples were let to swell until
equilibrium, which resulted in the materials having slightly different
solid contents during further measurements. A schematic representa-
tion of the matrix preparation and cell immobilization (Figure S1), a
table containing the composition of the cell immobilization matrices
before swelling (Table S1), and AFM topography and phase contrast
images of Ca2+-cross-linked TCNF and alginate films (Figure S2) are
shown in the Supporting Information.
2.5.1. Preparation of Matrices with TCNF. Matrices with TCNF
were prepared both with and without Ca2+-cross-linking. Materials
with calcium cross-linking were referred to with “Ca-0.5TCNF-
0.5ALG”, “Ca-0.9TCNF-0.1ALG”, “Ca-1.0TCNF”, and “Ca-
1.0TCNF-0.1PVA”, where the numbers before the main matrix
components, TCNF, and alginate accounted for their ratio in the final
1 wt % mixture. If present, PVA was added by an amount
corresponding to 10% of the dry weight of TCNF in the mixture,
which was signified with “0.1” before PVA in the name of the material.
Similarly, materials without calcium cross-linking were referred to
with “1.0TCNF” and “1.0TCNF-0.1PVA”. All matrices with TCNF
were prepared in three main steps: (i) mixing, (ii) casting, and (iii)
rewetting. During the mixing phase, matrix components were
homogenized with a digital T 25 Ultra-Turrax homogenizer (IKA,
Staufen, Germany) at 12,000 rpm for 3 min, after which air bubbles
were removed via centrifugation at 4500g for 3 min. In the mixing
phase, the material was prepared to contain 0.5 wt % of the main
matrix components (TCNF or TCNF-alginate combinations). Then,
the matrices were cast on a solid support coated with a PTFE film to a
thickness of 2 mm. Casting of the matrices was followed by the
primary cross-linking with Ca2+-ions. Thorough cross-linking of the
matrix materials was induced by spraying 50 mM CaCl2 on the gels
until wet, as described earlier.11,29 The gels without Ca2+-cross-linking
were sprayed with Milli-Q water until wet. Then, the matrices were
dewatered to a 1 wt % solid content in 23 °C and 50% relative
humidity (RH) to enhance penetration of Ca2+ to the nanofibril
network and to promote esterification between TCNF and PVA for
additional cross-linking. After dewatering, individual samples were cut
from the prepared materials. The samples with Ca2+-cross-linking
were then submerged in 50 mM CaCl2 for 15−30 min to eliminate
the chance of network collapse or loosening of the cross-link in freshly
cut sample edges. Finally, the samples were left to swell overnight
(16−24 h) before any successive measurements. Matrices without
cells were swelled in Milli-Q water, whereas matrices with
immobilized cells were swelled in the BG-11 growth medium.
2.5.2. Preparation of the Ca2+-Alginate Matrix. Calcium-alginate
matrices, referred to as “Ca-1.0ALG”, were otherwise treated similarly
to the matrices with TCNF in the mixing phase, but they were
prepared directly to a 1 wt % solid content. In the casting phase, the
mixture was poured on a solid support coated with a PTFE film to a
thickness of 1 mm and cross-linked by spraying 50 mM CaCl2 on top
until wetted. The Ca2+-ions were let to diffuse into alginate for
approximately 45 min, after which the samples were cut from the
solidified gel and rewetted as described for TCNF-containing
matrices.
2.5.3. Cell Immobilization. Wild-type Synechocystis cells were
immobilized to the prepared hydrogel matrices via passive gel
entrapment. Prior to immobilization, photosynthetically active
Synechocystis cells were collected by pelleting the cells by
centrifugation at 10,000g for 15 min. The cell concentrate was
diluted to have an optical density (OD720) of 1.0 ± 0.1. The optical
density measurements were performed with an AquaPen-C AP-C 100
handheld pulse amplitude modulation (PAM) chlorophyll fluorom-
eter (Photon Systems Instruments, Czech Republic) calibrated with
the BG-11 medium. The cell suspension was then added to the
hydrogel constituents in a 1:1 ratio during the mixing phase instead of
water or the BG-11 medium.
2.6. Rheological Mapping of Matrix Materials. 2.6.1. Small
Deformation Oscillatory Rheology. Thin-layer hydrogels with a flat
design were studied with small deformation oscillation stress sweep
measurements utilizing Discovery HR-2 and AR-G2 -rheometers (TA
Instruments, New Castle, DE, USA). Serrated parallel plate geometry
with a diameter of 40 mm was used in the measurements to avoid wall
slip. The serrated baseplate of the instrument was set at 22 °C with a
Peltier plate. Before measurements, the hydrogel samples were let to
adjust to compression for 2 min under the measuring head. The
normal force between the sample and the rheometer was followed
when lowering the measuring head to the thickness of the sample.
The gap was adjusted until a normal force between 0.3 and 1 N was
observed. Samples with a uniform thickness of 1000 ± 500 μm were
selected for further analysis, except for hydrogels without Ca2+-cross-
linking where all samples were thicker than 1500 μm after swelling.
Biomacromolecules pubs.acs.org/Biomac Article
https://doi.org/10.1021/acs.biomac.3c00261
Biomacromolecules 2023, 24, 3484−3497
3486
The stress sweeps were performed in an oscillation stress range from
0.1 to 600 Pa under a constant frequency of 0.1 Hz. During the
measurement, 10 points per measuring decade were taken.
Controlling the rheometer and data collection was conducted with
a TRIOS program (TA Instruments, New Castle, DE, USA).
With small deformation oscillatory rheology, the viscoelastic
properties of prepared hydrogels were interpreted in the linear
viscoelastic region (LVE region) and during material yielding. The
elastic and viscous components of a material were characterized by
the storage and loss moduli (G′ and G″, unit Pa).50 The loss tangent
(tan δ), or damping factor, was determined as the ratio of G″ to G′
(tan δ = G″/G′). G′, G″, and tan δ were determined from the linear
viscoelastic region (LVE region) of a stress sweep curve (Figure 2A),
at 1 Pa oscillation stress where G′ and G″ were independent of the
applied shear stress. Material yielding properties were characterized
with yield stress (σy) and critical stress (σc). Yield stress (σy) denoted
the onset of non-linear behavior in the modulus response to the
applied shear stress at the limit of the LVE region. It was determined
from the stress sweep curve as the point where G′ had decreased
approximately 5% from its preceding linear value (Figure 2A). Critical
stress (σc) indicated the system’s transition to predominantly
irreversible deformation. It was determined from the stress/strain
curve (Figure 2B) by obtaining the oscillation stress value where the
slope of the stress/strain curve had decreased 15−20% from the linear
region, i.e., the strain was no longer linearly dependent on the stress.
2.6.2. Multivariate Data Interpretation. Differences in the
rheological parameters of the hydrogel matrices were determined
using principal component analysis (PCA) and hierarchical clustering.
The measured rheological parameters (G′, G″, tan δ, σy, and σc) of the
samples were compiled into a 46 × 5 raw rheological data matrix
(Table S2) where the samples described in Section 2.5 were given as
row objects and the parameters as the corresponding columns. The
data matrix was preprocessed by normalizing the columns to unit
variance and zero mean to enable comparing variables given in
different units. The preprocessed data were then decomposed into
principal component (PC) scores and loadings according to the
general PCA model51 (eq 1):
= +
=
X tp E
i
n
nm
1
T
(1)
where Xm denoted the preprocessed and mean-centered data matrix, t
is the orthogonal score vectors, pT is the orthonormal loading vectors,
and En is the residual matrix after the calculation of n components.
With normalized and mean-centered data, the PC loadings were equal
to the eigenvectors associated with the largest eigenvalues of the
correlation matrix of Xm. The number of rheologically relevant
components was estimated based on the scree plot (Figure S3),52 and
the hydrogel differences were explained by the scores of each
component. The samples were then divided into distinct classes by
clustering the determined PC scores. The clusters were formed by
minimizing the sums of squared Euclidean distances between the
sample scores and the cluster centroids using Ward’s method.53 The
final number of clusters was determined by identifying increasing
cluster distances from the resulting dendrogram.43 Origin 2021b
(OriginLab Corporation, Northampton, MA, USA) was used to
calculate and plot the outputs from multivariate data analyses.
2.7. Thermoporosimetry. The pore structure of the hydrogels
was analyzed using the thermoporosimetry (TPM) method. The
measurements were conducted on a Mettler Toledo DSC 3+
(Mettler-Toledo Intl. Inc. Instrument, USA) differential scanning
calorimeter equipped with an intracooler. The samples in triplicates
were hermetically sealed in 40 μL aluminum pans. The masses of the
sealed crucibles were recorded before and after the measurements to
ensure that there was no leakage during the measurement. After the
measurements, the crucibles were punched with a needle and dried in
an oven at 105 °C overnight to determine the moisture content. The
temperature was first brought to −50 °C at 20 K min−1 to crystallize
all the freezable water in the samples. The temperature was then
increased to −0.2 °C and held constant until the melting transition
was completed, i.e., until all the water in the small capillaries melt.
This step is essential to prevent supercooling during the subsequent
recrystallization step. The temperature was then decreased at 2 K
min−1 to −50 °C. The relationship between the pore diameter (D)
and the melting temperature depression was described by the Gibbs−
Thomson equation, and the diameter was calculated from the
modified version according to Maloney.45 The pore volume
distributions in this study included only the fraction of water in the
sample that freezes. The interfacial water layer, which does not freeze,
represented only a small fraction of the overall pore volume and was
not included in the analysis. The pore volumes were expressed in
milliliters of water/g solids. It was assumed that the specific gravity of
water is 1, and deviations from this value due to temperature effects
were ignored.
2.8. SEM Imaging. Scanning electron microscopy (SEM) was
used to compare the network morphology in different matrices.
Hydrogel samples were cut into 5 mm × 5 mm pieces, and water was
removed through solvent exchange with ascending series of 30 × 100,
50 × 100, 70 × 100, 90 × 100, and 6 × 100% ethanol. The sample
was kept in each solution for 15 min, except for the first five 100%
ethanol steps that were 1 h each and the final exchange that occurred
overnight. Dehydrated samples were placed in a Bal-Tec CPD 030
critical point dryer (Bal-Tec, Liechtenstein) where ethanol was
substituted for liquid CO2 over several fill-purge cycles at 5−10 °C.
Finally, the temperature was increased to 40 °C to raise the pressure
to approximately 90 bar and bring CO2 above its supercritical point.
Gaseous CO2 was released, and the dried sample was mounted onto
an SEM stub with adhesive carbon tape. The sample was immediately
coated with 4 nm iridium in a Leica EM ACE600 sputter coater
(Leica Microsystems, Germany). SEM images were acquired using a
scanning electron microscope ZEISS Sigma VP with Gemini column
(ZEISS, Germany) at 1 kV acceleration voltage.
2.9. Gas Exchange Analysis from Immobilized Cells. An in-
house built membrane inlet mass spectrometry (MIMS) system54 was
used to monitor the in vivo O2 and CO2 fluxes from Synechocystis cells
entrapped in hydrogel matrices. In the technique, a semi-permeable
membrane in the MIMS chamber separates the sample from a mass
spectrometer, where ionized gases are detected and distinguished
based on their mass/charge (m/z) ratio.55 A 1 cm × 1 cm piece of the
Figure 2. Rheological parameters used as the basis for principal component analysis and rheological comparisons. (A) Typical stress curve of the
materials, where G′, G″, and tan δ (G″/G′) were determined from the linear viscoelastic (LVE) region at an oscillation stress of 1 Pa and σy from
the point where G′ decreases 5% from the LVE region. (B) Typical stress−strain curve of the materials, where σc was determined at the point of a
15−20% decrease from the linear slope.
Biomacromolecules pubs.acs.org/Biomac Article
https://doi.org/10.1021/acs.biomac.3c00261
Biomacromolecules 2023, 24, 3484−3497
3487
hydrogel matrix was placed in the MIMS sample chamber with 1 mL
of the BG-11 medium (pH 7.5). HCO3− was added to a final
concentration of 1.5 mM to ensure sufficient carbon supply for
photosynthesis during the experiments. The gas fluxes were then
monitored during a 5 min dark adaptation, a 5 min illumination
period of 500 μmol photons m−2 s−1, and a 5 min post-illumination
dark period. Gas exchange rates were calculated based on the
chlorophyll content in the hydrogel matrix pieces according to the
protocol reported by Beckmann et al.56 Maximal O2 (m/z 32)
evolution and CO2 (m/z 44) fixation rates were calculated as average
rates for 3−5 min after the onset of illumination and dark respiration
rates as average rates during pre-illumination darkness. Net O2
evolution was calculated as maximal O2 evolution − dark respiration.
2.10. Determination of the Chlorophyll Content in Hydro-
gel Matrices. The chlorophyll a (Chl) concentration was
determined by incubating the film pieces (1 cm × 1 cm) in 3 mL
of 90% (v/v) methanol at +4 °C overnight in the dark. Prior to
measurements, the samples were centrifuged with a tabletop
centrifuge with full speed for 1 min. Absorbance was measured
from the supernatant at 665 and 730 nm using a UV-1800
spectrophotometer (Shimadzu, Japan), and the 730 nm absorbance
values were subtracted from the 665 readings and multiplied by 12.7
to get the chlorophyll concentration.57
3. RESULTS AND DISCUSSION
3.1. Mapping the Rheological Properties of Hydrogel
Immobilization Matrices. We studied the rheological
properties of water-swollen hydrogel cell immobilization
matrices with different material constituents and cross-linkers
using small deformation oscillatory stress sweeps. The aim was
to determine systematic differences and potential groupings
within the materials based on their wet strength and visualize
these groups as a “rheological map”. Seven different hydrogel
matrices were prepared in water and without immobilized cells
using either alginate, TCNF, or their mixtures. The cross-
linking effect of Ca2+-ions was investigated with all matrices,
and PVA was used for TCNF-based matrices.
The PCA results described the rheological features of the
matrices and are shown in Figure 3A. The score plot illustrates
the differences within the samples, and the corresponding
loadings provide their interpretation based on the determined
rheological parameters. In PCA, positive score values have a
positive correlation with positive loadings and vice versa.
Because the PCs reduce multivariate measurement space to a
few interpretable dimensions, they can be used to explore and
identify important systematic variations in the data and are
useful for identifying sample groupings. As shown in Figure 3A,
the first two PCs explained 94% of the variation in the
preprocessed data, providing a meaningful interpretation of the
rheological properties of the hydrogels. The later PCs did not
provide any further information and as suggested by the scree
plot (Figure S3) were most likely attributed to measurement
noise.
Figure 3. Rheological map and hierarchical clustering of the matrix properties. (A) Rheological map based on the determined PCA scores (left)
and loadings (right). (B) Unsupervised hierarchical clustering of matrix materials presented as a dendrogram (left). Results from the clustering are
illustrated on the rheological map according to their assigned classes (right).
Biomacromolecules pubs.acs.org/Biomac Article
https://doi.org/10.1021/acs.biomac.3c00261
Biomacromolecules 2023, 24, 3484−3497
3488
The PCA illustrates that the TCNF-based matrices with
Ca2+-cross-linking (Ca-0.9TCNF-0.1ALG, Ca-1.0TCNF, and
Ca-1.0TCNF-0.1PVA) had high scores on PC1 and low scores
on PC2. Furthermore, the TCNF-based matrices without Ca2+-
cross-linking (1.0TCNF and 1.0TCNF-0.1PVA) had com-
parable scores on PC2 but negative scores on PC1. PC1
explained 58% of the data variation, and as seen from the
loadings, higher scores were associated with higher G′, G″, σy,
and σc and lower tan δ values compared to other matrices.
Thus, PC1 scores can be interpreted as the yield behavior and
wet strength of the hydrogels, and the loadings generally
suggested that elastic behavior at rest (i.e., input energy stored
within the material structure) led to later yielding and
progression to viscous behavior.
The matrices with a high alginate content (Ca-0.5TCNF-
0.5ALG and Ca-1.0ALG) were clearly different from the
TCNF-based matrices, possessing lower PC1 and higher PC2
scores. PC2 explained 36% of the remaining data variation, and
the corresponding loadings indicated that high scores were
associated with higher G′, G″, and tan δ values and, to some
extent, lower σy and σc values compared to other matrices. This
can be seen to reflect on the differences in the chemical
composition of the hydrogels and the cross-linking effect of
calcium ions before yielding. PC2 also suggested that
increasing the G″/G′ ratio, i.e., higher tan δ values, led to
material yielding in lower oscillation stresses during shearing.
The PCA sample scores were clustered into distinct classes
to assess if the rheological differences were indeed related to
specific matrix formulations. The scores of the first two PCs
were used, and the clusters were determined with Ward’s
method.53 The results are shown in Figure 3B. Three final
hydrogel classes were chosen based on increasing cluster
distances shown in the dendrogram. The first class included all
Ca-1.0ALG samples and three Ca-0.5TCNF-0.5ALG samples
(10 samples) with the highest rest behavior properties and was
therefore denoted as the “Ca-alginate class”. The second class
included two of the remaining Ca-0.5TCNF-0.5ALG samples
and all TCNF-based matrices with Ca2+-cross-linking (26
samples) and was named the “Ca-TCNF class”. Finally, the
remaining TCNF-based matrices without Ca2+-cross-linking
(10 samples) formed the “no calcium class”. Notably, the Ca-
TCNF class bore more resemblance to the Ca-alginate class
than to the no calcium class. Thus, it is clear that ionic cross-
linking in particular affected the rheological behavior of
TCNF-based hydrogel matrices.
The average rheological characteristics of each class are
shown in Table 1. With G′ values of over 1000 Pa, both Ca-
alginate and Ca-TCNF classes showed predominantly elastic
behavior and hence good capability of storing energy in their
internal chemical and physical bonds in the LVE region. The
matrices in the Ca-alginate class had especially high G′, which
lends rigidity to the material structures at rest. However, they
started yielding and progressed into predominantly viscous
deformation at oscillation stress values 2 times lower than the
Ca-TCNF class.
Ca2+-cross-linked matrices in the Ca-alginate and Ca-TCNF
classes were chosen for further rheological evaluations with
immobilized Synechocystis cells. The results at rest and during
yielding are shown in Figure 4.
With the cells included, the G′ and G″ values of all TCNF-
containing matrices fluctuated roughly around 1000 and 100
Pa, respectively. For Ca-1.0ALG matrices, the values were
much higher: G′ varied between 3000 and 5000 Pa and G″ at
300 Pa. The tan δ values increased with the proportional
alginate content, but here, Ca-0.5TCNF0.5ALG was com-
parable to Ca-1.0ALG, suggesting that it shares properties with
Table 1. Average (Mean ± SD) Rheological Properties of the Determined Hydrogel Classesa
rest properties yield properties
class G′ (Pa) G″ (Pa) tan δ (G″/G′) σy (Pa) σc (Pa)
Ca-alginate class 2400 ± 660 280 ± 86 0.12 ± 0.014 18 ± 14 52 ± 20
Ca-TCNF class 1500 ± 340 100 ± 17 0.071 ± 0.013 43 ± 18 100 ± 42
no calcium class 18 ± 4.0 2.0 ± 0.20 0.11 ± 0.016 2.0 ± 0.50 6.0 ± 3.0
aG′ = storage modulus, G″ = loss modulus, tan δ = the loss tangent (i.e., damping factor), σy = yield stress, σc = critical stress. Variables are defined
in Section 2.6. The values are rounded to two significant figures.
Figure 4. Rheological properties of Ca2+-cross-linked matrices from Ca-TCNF and Ca-alginate classes with immobilized Synechocystis cells. (A)
Materials at rest, showing G′, G″, and tan δ. (B) Materials during yielding, showing yield stress (σy) and critical stress (σc). Sample size: n = 4 for
Ca-1.0TCNF and Ca-1.0ALG, n = 6 for Ca-0.9TCNF0.1ALG and Ca-0.5TCNF0.5ALG, and n = 11 for Ca-1.0TCNF0.1PVA.
Biomacromolecules pubs.acs.org/Biomac Article
https://doi.org/10.1021/acs.biomac.3c00261
Biomacromolecules 2023, 24, 3484−3497
3489
both matrix types. As with the PCA-assisted mapping, the
results reflected the properties of the dominating matrix
component, i.e., the energy lost to dissipation (viscous
behavior) versus stored in the material (elastic behavior)
increased in the matrices with higher alginate contents.
The ratio of elastic to viscous behavior of the matrices at rest
seems also to correlate with the yield properties shown in
Figure 4B. Ca-1.0TCNF had the highest and Ca-1.0ALG the
lowest yield properties with the mixed matrix formulations in
between. However, here, the difference between Ca-1.0TCNF
and the other TCNF-containing matrices was much larger.
Moreover, Ca-1.0TCNF-0.1PVA and Ca-0.9TCNF-0.1ALG
had very similar rheological properties overall, which suggests
that a small addition of either polymer softens the TCNF
matrix in a comparable manner. The ratio of yield stress to
critical stress was consistent throughout the different matrices,
except for Ca-1.0ALG, which began to yield at much lower
critical stress than any other material. Finally, the addition of
cells caused minor changes in the rest properties of all matrices
(Figure S4), with the tan δ values of TCNF-containing
matrices increasing slightly. Yield values, however, fall well
within the error margins for all samples.
Determining the wet mechanical properties of the hydrogel
matrices subject of this investigation is challenging due to the
unique nature of their structures. The samples are not strong
enough for conventional tensile testing, and results from
compression measurements do not represent real conditions in
photobioreactors very well. In contrast, shear stresses are
reported to have a particularly profound effect in these
systems.58−60 The effect of shear stress on materials can be
studied with small and large deformation oscillatory rheology.
The former gives accurate information in the LVE region and
at the onset of nonlinear (yielding) behavior of gel-like
materials,61 while the latter provides insight to their flow and
fracture properties.62 This makes small deformation rheology
more favorable for our investigations as we are not interested
in the flow properties of the materials, and the formation of
fractures can be deemed more indicative of heterogeneous
structural weaknesses in a gel matrix in comparison to yield
properties.63
Nonetheless, we have found that the matrices investigated
here are inherently heterogeneous and also much stiffer than
typical samples measured with small deformation oscillatory
rheology, which leads to high variance, especially after the
onset of nonlinear behavior. This is largely due to the increased
likelihood of slipping as the shape and roughness of the sample
surfaces are difficult to control, leading to uneven contact in
the plate−plate measurement setup. The use of a serrated plate
is known to alleviate these difficulties,64 and we have improved
the reproducibility through a systematic approach in sample
preparation, validation, and measurement protocols presented
here. For further in-depth mechanical investigations of self-
standing hydrogels, we also recommend using localized stress
sweeps with added measurement points in the predetermined
yield region of the materials to improve the measurement
sensitivity in the region.
Here, we studied the rheological properties of the samples
using a dataset with higher number of replicates than in our
previous study11 and identified clear trends across different
matrix compositions based on PCA and clustering. The results
were presented as a rheological map, which successfully
differentiated the materials into distinct classes (Figure 3). Our
approach is novel in studying the rheology of self-standing
matrices and offers two main advantages. First, PCA rotated
the original measurement axes toward the direction of
maximum variation and captured the most important
correlations across the rheological measurements in the PCA
loadings. These loadings provide an interpretation of the
dominant rheological features of the matrices by separating
non-systematic variations generated by empirical uncertainties.
Second, the corresponding PCA scores enabled us to visualize
the most important differences across the samples in just a few
interpretable dimensions. We further clustered these scores
into distinct classes, which describe the main matrix types
based on their determined rheological characteristics.
According to the results, the hydrogel matrices yielded at
higher oscillation shear stresses (high σy and σc) with
increasing TCNF concentration, although they possessed
weaker rest behavior properties than the matrices with high
alginate concentrations. Notably, without the ionic cross-
linking, the TCNF-based matrices had weak rest and yield
properties and were not self-standing, even with the
incorporation of PVA as an additional cross-linker. With
increasing alginate concentration, the Ca2+-cross-linked ma-
trices became more rigid with high elastic capacity at rest, but
due to their simultaneous susceptibility to energy dissipation,
they yielded in low oscillation stresses.
The dual nature of high G′ and low yield values for alginate-
based matrices is likely a result from intrinsic attributes of the
chemistry behind the gelation of alginate. Alginate forms gels
with divalent cations through development of specific
coordination complexes that are often explained via the “egg-
box model”, proposed by Grant et al.65 Although the egg-box
model has undergone revisions during the 21st century,66,67 it
serves as the principally accepted mechanism and foundation
for the multistep gelation process of alginate.68,69 The gelation
is understood to be a synergistic effect of intramolecular
stereochemistry, polyelectrolyte effect, and sequential associa-
tion between alginate chains: the Ca2+-ions bond with the
deprotonated carboxylate anions present in G residues in a
critical nucleation step followed by the formation of egg-box
dimers between successive G units (GG-blocks) and their
lateral association to multimers.68,69
The ionic cross-links in alginate are reversed if stronger
ligands than those participating in the gelation process are
added to the system. This causes an alginate gel to dissolve
into a viscous liquid. Similarly, if high enough input stress is
introduced to the gel, then the egg-box structures begin to
break, or “unzip”, and irreversible material deformation via
fracture initiation and propagation takes place.70 Thus,
alginate-based immobilization matrices are inherently suscep-
tible to both chemical and mechanical stresses. Furthermore,
alginate contains MM-blocks and mixed MG-blocks in ratios
depending on the source of the alginate. The blocks that
contain mannuronic acid subunits interact weakly with Ca2+-
ions in comparison to GG-blocks, effectively reducing the
active surfaces required for gelating.67 Overall, the ion-
coordination complexes produce elastic capacity in alginate
matrices, leading to a more rigid material at rest, but the
reversibility and specificity of the bonds along with the viscous
nature of alginate solutions make alginate-based gel matrices
susceptible to breaking under shear stress.
Compared to alginate, TCNF are larger structural units
consisting of numerous cellulose polymer chains with
carboxylic groups only on the C6 positions. Thus, the
interfibril Ca2+-cross-links of TCNF via Ca2+-carboxylate
Biomacromolecules pubs.acs.org/Biomac Article
https://doi.org/10.1021/acs.biomac.3c00261
Biomacromolecules 2023, 24, 3484−3497
3490
complexes71 must bind much larger structures together than
the interchain Ca2+-cross-links between individual alginate
polymer chains.5,67,69 However, due to the high aspect ratio of
TCNF, they form a highly viscous gel-like material at a low
solid content even without cross-linking. Upon addition of
divalent or multivalent cations, the interfibril repulsive forces
caused by the carboxylate surfaces are effectively screened71
and the formation of Ca2+-carboxylate complexes, though
weaker than in the egg-box structures of alginate, increases
their overall yield and critical stresses to higher values than
what has been measured for alginate hydrogels. Last, adding to
the cross-linking effect of Ca2+-ions, the entangled colloidal
fibril network of TCNF can resist deformation effectively once
the yielding starts. One suggested mechanism for this is the
presence of contiguous structures, i.e., a continuous system of
connections with “structural viscosity”, within the nano-
cellulose networks,61 which may also allow the formation of
new Ca2+-carboxylate complexes even during deformation,
creating a so-called “self-healing network” when TCNF transfer
deformative load over the length of the fibrils.
The addition of PVA, which has previously been reported to
act as a cross-linking agent in TCNF films,41 and suggested to
aid in the cross-linking of TCNF hydrogels in our previous
efforts,11,34 appears to have a net negative impact on the yield
properties when compared to pure Ca2+-TNCF matrices. The
TCNF-based matrices also had very weak mechanical proper-
ties without Ca2+-cross-linking, both with and without PVA.
These results suggest that the ionic complexation between
Ca2+ and TCNF is the primary cross-linking element in these
matrices and that the TCNF-PVA cross-link would seem to
require almost complete dewatering of the TCNF hydrogel for
it to take place via esterification as reported earlier.41 The
addition of anionic (alginate) or non-ionic (PVA) polymers
into TCNF suspensions could induce entropic depletion-
induced flocculating or adsorption and subsequent polymer
bridging,64 but here, these effects seemed to be negated by the
disruption of the contiguous TCNF network and ionic
coordination.
Finally, we observed only a minor change in the wet
mechanical properties of the matrices by the inclusion of
entrapped Synechocystis cells. In our previous effort, we
reported a decrease in yield properties for both TCNF- and
alginate-based matrices using the same biomass loadings,11 but
our larger dataset here seems to suggest that these changes are
within the error margins. However, this is likely to change if
higher biomass loadings are used and is a subject for future
efforts.
3.2. Comparing the Porous Structures of Ca2+-Cross-
Linked Matrices. We used TPM to directly measure the
mesoporosity and pore size distribution of the matrix hydrogels
Figure 5. Porosity of the matrices. (A) Cumulative and (B) derivative pore size distribution of the water-swollen cross-linked matrices measured via
TPM and (C) SEM images of the cross-linked hydrogel matrices prepared via critical point drying as well as a photo showing the appearance of a
water-swollen Ca-1.0ALG hydrogel.
Biomacromolecules pubs.acs.org/Biomac Article
https://doi.org/10.1021/acs.biomac.3c00261
Biomacromolecules 2023, 24, 3484−3497
3491
in the wet state. Figure 5A shows the cumulative and Figure 5B
the differential pore size distribution (PSD) of the matrices
after rinsing and swelling in water for over 48 h. In addition, we
used SEM imaging of the same matrices (Figure 5C), prepared
via critical point drying, to gain insight on the morphology and
macroporosity of the materials.
As seen in Figure 5A, the cumulative pore volumes of the
matrices increase in a heterogeneous and hierarchical manner,
as indicated by the onset of a secondary slope at a pore size of
approximately 20 nm. The largest total pore volume is
observed for Ca-1.0TCNF-0.1PVA followed by 0.9TCNF-
0.1ALG (Table 2), both of which have higher porosity than the
other matrices, especially below the 20 nm pore size. Below
them are Ca-1.0TCNF and Ca-0.5TCNF-0.5ALG with very
similar pore volumes followed by 1.0ALG with a much lower
pore volume than the other matrices. Thus, TCNF-based
matrices were observed to have higher mesoporosity than pure
Ca-alginate, but interestingly, a small addition of soluble PVA
or alginate polymers to the TCNF network decreased its pore
size but increased its total (meso)pore volume. In the SEM
images (Figure 5C), all TCNF-containing matrices, including
Ca-0.5TCNF-0.5ALG, seemed to share the typical appearance
of a fibrillar network with larger pores than pure Ca-1.0ALG,
which appears to have a somewhat denser network structure
with much less apparent porosity. However, pure Ca-1.0TCNF
seems to have the highest number of large pores based on the
images. This can also be seen in the PSD in Figure 5B, where
Ca-1.0TCNF peaks at the highest pore size at ∼50 nm and Ca-
1.0ALG at the lowest at 30 nm. Ca-1.0TCNF-PVA also has a
sharp peak at ca. 30 nm, and the peaks for Ca-0.9TCNF-
0.1ALG and Ca-0.5TCNF-0.5ALG place in between the others
at around 35 and 40 nm, respectively. The volume-weighed
mean pore sizes shown in Table 2 correlate well with the PSD
peaks, with the exception of Ca-1.0ALG, whose second highest
mean pore size is likely due to its low pore volume at sizes
below 20 nm.
As expected, the water contents of the swollen matrices
shown in Table 2 have increased from the original 99 wt %
after matrix preparation. The primarily TCNF-based matrices
have a water content of approximately 99.3 wt %, and the
proportional addition of alginate decreases it moderately to
approximately 99.1 wt % of Ca-1.0ALG, indicating a lower
amount of swelling and thus lower overall porosity after the 48
h immersion in water.
Thermoporosimetry is used to characterize ridged meso-
porous materials such as silica gels.45 It has also been
successfully applied for swollen but largely insoluble materials
such as cellulosic fibers, and the evidence to date is largely
supporting the validity of this technique.45 Moreover, TPM
also appears to give useful information for hydrated polymeric
materials that are largely dissolved, such as alginate and PVA.
However, for this material class, it is unclear if the usual
Gibbs−Thomson coefficient is correct and if other effects such
as ice crystal damage cause serious artifacts.
In the present study, we apply TPM to an even more
challenging system: water-swollen complex multicomponent
gels with colloidal TCNF and polymers that can be partially
dissolved and partially cross-linked (PVA, alginate). The
details of the water association in such systems are very
complex and difficult to analyze. The organization of the
polymers, localized steric effects, competitive hydration, and
other effects is expected to affect the binding of water within
the matrix. Nonetheless, the TPM results presented here seem
to provide some insights on the porosity of the swollen gels.
Another indication pointing toward the reliability of the
measurement is that the total pore volume of Ca-1.0TCNF is
very similar to that of a pure TCNF hydrogel with the same
solid content measured using the same technique by Guccini et
al.,46 which suggests that the results are reproducible.
The main finding from the TPM data was the difference in
porosity between TCNF-containing and pure alginate
matrices. This is likely due to their capacity to swell in
water, which directly enlarges the cavities and loosens the
matrix network. The high hygroscopicity of TCNF allows the
network to hold more moisture and thus loosens the network
and enlarges the porosity of the fibril network.72 Moreover, the
different cross-linked structures could affect their morphology.
Alginate polymers are softer than TCNF, but the egg-box
structures of Ca2+-cross-linked alginate provide a relatively
rigid network with some steric limitations in its orientations
and much reduced electrostatic repulsion. This allows alginate
to form tight and uniform mesoporous networks. On the other
hand, Ca2+-TCNF cross-linking is sparser and allows the fibril
network more freedom to swell, even if the TCNF network is
rigid by itself.
The onset of a secondary slope in the cumulative PSD seen
in all samples (Figure 5A) indicates the formation of
hierarchical pore structures, as also discussed by Guccini et
al.46 The largest heterogeneous behavior is observed with Ca-
1.0TCNF-PVA, possibly due to PVA forming secondary
structures inside the hydrogels. This is supported by the
rheology measurements, indicating that TCNF and PVA are
not directly attached to one another in the system. Similarly,
the alginate in Ca-0.9TCNF-0.1ALG and Ca-0.5TCNF-
0.5ALG could form heterogeneous networks with secondary
structures, which would explain their relatively high pore
volumes compared to pure Ca2+-cross-linked TCNF, while
pure Ca2+-cross-linked alginate has a much less porous
structure.
It is important to note that TPM is limited to detect the
mesoporosity (2−50 nm) of a given material. Hence, we
investigated SEM images of the same samples obtained via
critical point drying (Figure 5C) to gain complementary
insight on the morphology and macroporosity of the materials.
This method avoids the crossing of phase transition
boundaries, preventing the collapse of the fibrillar network
more effectively than other SEM sample preparation methods
such as air- or freeze-drying.73,74 Thus, even if the interactions
between fibril and polymer chains after solvent exchange might
not fully represent the interactions in the water-swollen
matrix,36,75,76 qualitative insight of the matrix microstructure
and macroporosity can be gained. The clearest observation
Table 2. Solid Content, Volume Weighed Average Pore
Size, and Total Pore Volume of the Swollen Matrices
matrix
composition
solid content
(wt %)
mean pore size
(nm)
total pore volume
(mL/g)
Ca-1.0TCNF 0.71 ± 0.010 52 ± 3.0 16 ± 0.44
Ca-1.0TCNF-
PVA
0.70 ± 0.14 42 ± 0.44 24 ± 0.47
Ca-0.9TCNF-
0.1ALG
0.71 ± 0.15 42 ± 0.12 20 ± 1.1
Ca-0.5TCNF-
0.5ALG
0.84 ± 0.060 40 ± 0.47 15 ± 0.59
Ca-1.0ALG 0.91 ± 0.12 47 ± 2.6 9.8 ± 0.37
Biomacromolecules pubs.acs.org/Biomac Article
https://doi.org/10.1021/acs.biomac.3c00261
Biomacromolecules 2023, 24, 3484−3497
3492
from the images was that the cross-linked alginate polymers in
Ca-1.0ALG create thin structures that form a more branched
and more densely arranged network with no observed
macroporosity, when compared to TCNF-containing matrices
containing distinctly visible colloidal fibrils and much larger
and more heterogeneously sized voids. This further highlights
the distinctions between the two materials and their cross-
linking pathways with Ca2+, supporting the main findings
gained from both rheology and TPM measurements.
Otherwise, all TCNF-containing matrices seemed to possess
a similar morphology with little visible distinctions besides a
slightly higher amount of macropores on Ca-1.0 TCNF. Thus,
the comparatively higher mesoporous volumes of Ca-
1.0TCNF-0.1PVA and Ca-0.9TCNF-0.1ALG were not re-
flected in their apparent macroscale porosity. This could
indicate that the suggested hierarchical structures within the
colloidal TCNF network are more localized in nature. Indeed,
their high pore volume especially at lower pore sizes below 20
nm (Figure 5A) combined with lower average pore size than
pure TCNF matrices (Figure 5B) suggests that the soluble
polymeric PVA and alginate can occupy the larger macropores
otherwise filled with free water and create mesoporous
structures within them, resulting in an increase in their
mesoporous volume while the amount of visible macropores
decreases. However, investigating these findings in more detail
is a topic for future efforts. All in all, it appears that the SEM
images are more in line with the PDS peaks in Figure 5B and
the average pore sizes of the materials reported in Table 2 than
the cumulative pore volumes shown in Figure 5A.
3.3. Investigation of Biocompatibility and Mass
Transfer Properties of Matrices with Immobilized
Cyanobacteria. We employed a highly sensitive MIMS
approach to monitor O2 and CO2 gas fluxes of wild-type
Synechocystis cells entrapped in the different hydrogel matrices
for 5 min of illumination and in darkness to examine their
photosynthetic electron transport and CO2 fixation capacities
(Figure S5). Maximal rates of O2 evolution and CO2 uptake
from the medium into the entrapped cells were reached after
ca. 3−5 min of illumination when CO2 fixation reactions
became activated. One day after cell immobilization, we
measured high net O2 evolution and CO2 fixation rates, at ca.
300 μmol O2 mg Chl−1 h−1 and 225−300 μmol CO2 mg Chl−1
h−1, respectively, from all hydrogel matrix types (Figure 6A,B).
Despite careful cell loading, fluctuations in the initial
chlorophyll contents in the matrices magnified the variation
in the measured gas fluxes. Still, we measured higher CO2
fixation rates from cells that immobilized Ca-1.0TCNF and
Ca-1.0TCNF-0.1PVA hydrogel matrices in comparison to
other matrix types (Figure 6B). Although these differences
were not statistically significant according to a two-way analysis
of variance (ANOVA), they seem to have a connection with
the higher pore size and volume observed, respectively, in
TCNF- and TCNF-PVA-based hydrogel matrices via thermo-
porosimetry and SEM imaging (Figure 5).
Figure 6. Rates of O2 evolution and CO2 uptake of cells entrapped in hydrogel matrices as measured by membrane inlet mass spectrometry.
Maximal rates of net O2 evolution (A) and CO2 fixation (B) measured at the steady-state phase of photosynthetic induction calculated as average
rates between 3 and 5 min after the onset of illumination. Values in panels (A, B) are means ± standard deviation from two (day 1) to three (days 3
and 5) biological replicates. (C) Photographs of the different hydrogel matrices with entrapped Synechocystis cells at 0, 3, and 5 days after
entrapment.
Biomacromolecules pubs.acs.org/Biomac Article
https://doi.org/10.1021/acs.biomac.3c00261
Biomacromolecules 2023, 24, 3484−3497
3493
Three days after immobilization, the chlorophyll contents
have stabilized to more equal values between matrix types, and
we measured net O2 evolution and CO2 fixation rates around
200 μmol O2 mg Chl −1 h−1 from all hydrogel matrices. No
significant differences were detected between different matrices
according to ANOVA (Figure 6A,B). Photosynthetic gas fluxes
decreased further on day 5 after immobilization, reaching rates
around 150 μmol O2 mg Chl−1 h−1 for net O2 evolution and
CO2 fixation. The decrease in photosynthetic performance
across all matrix types between day 1 and day 5 was statistically
significant according to ANOVA (p < 0.05). Moreover, the
most notable decrease was observed in the CO2 fixation
capacity of cells immobilized in the Ca-1.0TCNF hydrogel
matrix (Figure 6B). Photographs of the matrices (Figure 6C)
indicate that all samples appeared relatively similar throughout
the experiments, with slight differences in cell growth and
accumulation of bubbles inside the matrices, which are most
likely due to the heterogeneous pore structure of the samples.
However, one noticeable difference was that Ca-1.0ALG
showed less formation of macroscopic bubbles inside the
matrix, which is in line with the observations from the SEM
imaging that Ca-alginate has the tightest and most ordered
pore structure with no observed macroporosity.
MIMS is a versatile method utilized to investigate in vivo
exchange of gaseous compounds in cells or tissues, allowing
real-time, simultaneous monitoring of, e.g., photosynthetic
oxygen and carbon fluxes.55 Here, to our knowledge, a MIMS
setup was used for the first time to study the O2 evolution and
CO2 fixation in photosynthetic cells entrapped within thin
hydrogel films. The MIMS measurement showed that both
TCNF- and alginate-based immobilization matrices allow
efficient evolution of O2 as well as uptake and fixation of
CO2 (Figure 6 and Figure S5). Indeed, 3 days after
immobilization, all the immobilized systems show gas exchange
rates comparable to suspension cultures of Synechocystis.77,78
However, the reduction in the rates through days 1−5
indicates that the exchange is hindered over time, either due
to growth or biological changes in the immobilized cells or
some restriction in the mass transfer properties of the matrices.
The time-dependent differences between TCNF and other
matrices also highlight that both the cells and the matrices are
in a constant dynamic state in the SSPCF platforms.
Nonetheless, the results demonstrate that all immobilization
strategies were conducive to CO2 uptake from the medium
into the entrapped cells as well as the operation of the
photosynthetic electron transport chain that allow efficient
electron transport and carbon metabolism toward various
photosynthesis-based products. This makes Synechocystis cells
entrapped in all the self-standing hydrogel matrix compositions
attractive platforms for various photosynthesis-based biopro-
duction applications, depending on which material and
biological properties are most desirable.
When combined, our multidisciplinary analysis yielded some
intriguing connections between photosynthetic performance
and the distinct properties of the matrix materials. First, the
larger intrinsic pore size and volume of TCNF-containing
matrices in comparison to their alginate-based counterparts, as
shown via TPM and SEM imaging in Section 3.2, likely allow
more efficient photosynthetic CO2 exchange soon after
immobilization. In contrast, the air bubbles that appear in
TCNF-containing matrices after 3 and 5 days of cultivation
(Figure 6C) could indicate impairment of photosynthetic gas
exchange, as is also suggested by the notable decrease in
maximum CO2 fixation rate 5 days after immobilization in the
Ca-1.0TCNF matrix (Figure 6B). Our rheological analysis
showed that the Ca-1.0TCNF hydrogel matrix had a
contiguous structure with higher yield properties than other
matrices (Figure 4B). Combined with the thixotropic behavior
of TCNF, this may allow it to resist the loosening of the matrix
network due cell division and the reversal of the ionic cross-
linking due to ion exchange in the BG-11 medium11 more than
the other matrices over time. By day 5, this may result in
restricted gas exchange though the Ca-1.0TCNF matrix,
limiting the cells’ ability to fix CO2 in the Calvin−Benson−
Bassham cycle. In contrast, the initially more uniform and
denser alginate matrix does not seem to have macroscopic
bubble formation and maintains CO2 fixation performance
slightly better than the TCNF matrix over time, suggesting that
the high viscous behavior and low yield properties can facilitate
the gas exchange, albeit at the cost of mechanical stability, due
to higher tendency for swelling over time.
Further studies with larger sample sizes will be required for a
more in-depth elaboration on the interdependencies between
photosynthetic gas exchange and mass transfer or structural
properties of the different hydrogel matrices. Nevertheless, this
study provides a proof of concept demonstrating the
applicability of gas flux analysis in the interdisciplinary
development process of SSPCFs and providing means to
study the photosynthetic performance and mass transfer
limitations of the cells directly within immobilized systems.
It also provides detailed information about the biocompati-
bility of the cells with their immobilization matrices and could
aid in the estimation of changes in the biochemistry of the
entrapped cells in comparison to cells cultivated in suspension.
For example, combining MIMS measurements with absorb-
ance spectrometry would produce highly specific data about
the functionality of the whole photosynthetic electron
transport chain as well as the downstream metabolic reactions
of entrapped cells. Last, when combined with mechanical
(rheology) and structural (TPM and SEM) analyses of the
matrices, these biological performance indicators can be
directly connected to the dynamic structural and mechanical
properties of the matrices.
4. CONCLUSIONS
This study showcases an interdisciplinary toolbox of
experimental methods that demonstrates how the evaluation
of structure−property performance interfaces can assist in the
development of TCNF- and alginate-based SSPCFs. The
rheological data from the immobilization matrices without
cells, interpreted with PCA, highlighted the importance of
Ca2+-ions as the main cross-linking constituent. Moreover,
matrices with a high TCNF concentration yielded at higher
oscillation shear stresses (high σy and σc) but possessed weaker
rest behavior properties (G′, G″, and tan δ) than matrices with
high alginate concentrations. These mechanical properties,
which reflect the operational stability of the cell immobilization
matrices, can be explained through the innate structural
differences between the alginate polymer chains and colloidal
contiguous TCNF “self-healing” network, as well as their ionic
cross-linking mechanisms via egg-box coordination and Ca2+-
carboxylate complexation, respectively. The results indicate
that these properties can be controlled via altering, e.g., the G/
M ratio and block lengths of alginate and the charge and fibril
size of TCNF.
Biomacromolecules pubs.acs.org/Biomac Article
https://doi.org/10.1021/acs.biomac.3c00261
Biomacromolecules 2023, 24, 3484−3497
3494
For the determination of porosity, we demonstrated that
TPM can be utilized for complex water-swollen multi-
component gels with colloidal TCNF and partially dissolved
and partially cross-linked polymers (alginate, PVA). The
matrices form heterogeneous and hierarchical mesoporous
networks, especially in the presence of non-cross-linked
polymers (Ca-1.0 TCNF-PVA and Ca-0.9TCNF-0.1ALG). In
general, TCNF-based matrices were more porous than the
ones based on alginate. The difference in porosity is likely due
to different swelling capacities between TCNF and alginate.
SEM imaging supported the results from TPM in the
macroporous range as Ca2+-cross-linked TCNF matrices
appeared to form thicker networks with distinctly visible
colloidal fibrils and large voids, while Ca2+-cross-linked alginate
matrices formed visibly thinner and denser structures. Overall,
the porosity measurements provide a link into the interface
between the matrix structure and mass transfer properties.
The biological performance was studied with gas flux
analysis by utilizing MIMS, for the first time, for cells
entrapped within thin-layer hydrogel immobilization matrices.
In general, gas flux analysis with MIMS unveils the operational
status of the entrapped producer cells and, when combined
with porosity measurements, can also be used to study mass
transfer limitations across different immobilization systems. It
was shown that all matrix compositions investigated here
enabled efficient O2 evolution and CO2 fixation in entrapped
photosynthetic cells, with levels comparable to suspension
cultivations after 3 days of operation. Interestingly, we
observed that a differential decrease in photosynthetic
performance over time between hydrogel matrix types
appeared to connect with the porosity and yield properties
of the matrices. An interdisciplinary approach such as
employed here consisting of the rheology, porosity, and gas
exchange measurements can thus explain how the structure
and mechanical properties are dynamically linked to the
performance of immobilized photosynthetic cells. Moreover,
our approach revealed the interdependency of properties such
as porosity and mechanical stability: colloidal and thixotropic
TCNF create a strong network during transition from rest to
shear-induced deformation, and the TCNF-based matrices can
resist the loss of ionic cross-linking well due to their contiguous
and highly viscous nature. On the other hand, Ca-alginate
matrices, which are initially rigid but may yield more easily, can
offer improved gas exchange capabilities through additional
matrix loosening over time. Overall, these techniques offer a
way to interlink the mass transfer properties, porosity, and
operational stability in the SSPCF development. Ultimately,
the work shown here facilitates a pathway toward SSPCFs
tailored to the specific needs of the production organism and
process conditions.
■ ASSOCIATED CONTENT
*sı Supporting Information
The Supporting Information is available free of charge at
https://pubs.acs.org/doi/10.1021/acs.biomac.3c00261.
Graphical overview of cell immobilization procedures;
composition of the cell immobilization matrices before
swelling; AFM images of Ca2+-cross-linked TCNF and
alginate films; raw rheological data matrix used for PCA;
scree plot for PC selection; rheological properties of self-
standing matrices with and without Synechocystis cells;
fluxes of O2 evolution and CO2 uptake in entrapped cells
(PDF)
■ AUTHOR INFORMATION
Corresponding Authors
Ville Rissanen − VTT Technical Research Centre of Finland
Ltd., VTT, FI-02044 Espoo, Finland; orcid.org/0000-
0001-5750-8989; Email: ville.rissanen@vtt.fi
Tekla Tammelin − VTT Technical Research Centre of
Finland Ltd., VTT, FI-02044 Espoo, Finland; orcid.org/
0000-0002-3248-1801; Email: tekla.tammelin@vtt.fi
Authors
Tuukka Levä − VTT Technical Research Centre of Finland
Ltd., VTT, FI-02044 Espoo, Finland
Lauri Nikkanen − Molecular Plant Biology, Department of
Life Technologies, University of Turku, FI-20014 Turku,
Finland
Vilja Siitonen − Molecular Plant Biology, Department of Life
Technologies, University of Turku, FI-20014 Turku, Finland
Maria Heilala − Department of Applied Physics, Aalto
University, FI-00076 Espoo, Finland; orcid.org/0000-
0002-4237-4081
Josphat Phiri − Department of Bioproducts and Biosystems,
Aalto University, FI-00076 Espoo, Finland; orcid.org/
0000-0002-7445-5265
Thaddeus C. Maloney − Department of Bioproducts and
Biosystems, Aalto University, FI-00076 Espoo, Finland
Sergey Kosourov − Molecular Plant Biology, Department of
Life Technologies, University of Turku, FI-20014 Turku,
Finland; orcid.org/0000-0003-4025-8041
Yagut Allahverdiyeva − Molecular Plant Biology, Department
of Life Technologies, University of Turku, FI-20014 Turku,
Finland
Mikko Mäkelä − VTT Technical Research Centre of Finland
Ltd., VTT, FI-02044 Espoo, Finland; orcid.org/0000-
0001-6174-6330
Complete contact information is available at:
https://pubs.acs.org/10.1021/acs.biomac.3c00261
Author Contributions
⊥T.L., V.R., and L.N. contributed equally. The manuscript was
written through contributions of all authors. All authors have
given approval to the final version of the manuscript.
Funding
The work has been financially supported by the Academy of
Finland (AlgaLEAF, project nos. 322752, 322754, and
322755). This project has been partially financed by the
European Union’s Horizon 2020 research and innovation
program under grant agreement no. 899576 (FuturoLEAF).
T.L., V.R., M.M., T.T., T.C.M., and M.H. acknowledge the
Academy of Finland’s Flagship Programme under project nos.
318890 and 318891 (Competence Center for Materials
Bioeconomy, FinnCERES). Y.A. acknowledges the financial
support by NordForsk NCoE program “NordAqua” (project
82845). J.P. and T.C.M. appreciate the financial support from
the Foundation for Research of Natural Resources in Finland.
Notes
The authors declare no competing financial interest.
Biomacromolecules pubs.acs.org/Biomac Article
https://doi.org/10.1021/acs.biomac.3c00261
Biomacromolecules 2023, 24, 3484−3497
3495
■ ACKNOWLEDGMENTS
We acknowledge the provision of facilities and technical
support for SEM imaging by Aalto University at OtaNano -
Nanomicroscopy Center (Aalto-NMC). We are thankful to
Tuuli Virkkala for assistance in hydrogel matrix preparation.
■ REFERENCES
(1) Cascone, M. G.; Laus, M.; Ricci, D.; Sbarbati Del Guerra, R.
Evaluation of Poly(Vinyl Alcohol) Hydrogels as a Component of
Hybrid Artificial Tissues. J. Mater. Sci.: Mater. Med. 1995, 6, 71−75.
(2) Drury, J. L.; Mooney, D. J. Hydrogels for Tissue Engineering:
Scaffold Design Variables and Applications. Biomaterials 2003, 24,
4337−4351.
(3) Jen, A. C.; Wake, M. C.; Mikos, A. G. Review: Hydrogels for Cell
Immobilization. Biotechnol. Bioeng. 1996, 50, 357−364.
(4) Lee, J.; Cuddihy, M. J.; Kotov, N. A. Three-Dimensional Cell
Culture Matrices: State of the Art. Tissue Engineering - Part B: Reviews
2008, 14, 61−86.
(5) Andersen, T.; Auk-Emblem, P.; Dornish, M. 3D Cell Culture in
Alginate Hydrogels. Microarrays 2015, 4, 133−161.
(6) Leenen, E. J. T. M.; Dos Santos, V. A. P.; Grolle, K. C. F.;
Tramper, J.; Wijffels, R. H. Characteristics of and Selection Criteria
for Support Materials for Cell Immobilization in Wastewater
Treatment. Water Res. 1996, 30, 2985−2996.
(7) Mallick, N. Biotechnological Potential of Immobilized Algae for
Wastewater N, P and Metal Removal: A Review. BioMetals 2002, 15,
377−390.
(8) de-Bashan, L. E.; Bashan, Y. Immobilized Microalgae for
Removing Pollutants: Review of Practical Aspects. Bioresour. Technol.
2010, 101, 1611−1627.
(9) Léonard, A.; Dandoy, P.; Danloy, E.; Leroux, G.; Meunier, C. F.;
Rooke, J. C.; Su, B. L. Whole-Cell Based Hybrid Materials for Green
Energy Production, Environmental Remediation and Smart Cell-
Therapy. Chem. Soc. Rev. 2011, 40, 860−885.
(10) Vajravel, S.; Sirin, S.; Kosourov, S.; Allahverdiyeva, Y. Towards
Sustainable Ethylene Production with Cyanobacterial Artificial
Biofilms. Green Chem. 2020, 22, 6404−6414.
(11) Rissanen, V.; Vajravel, S.; Kosourov, S.; Arola, S.; Kontturi, E.;
Allahverdiyeva, Y.; Tammelin, T. Nanocellulose-Based Mechanically
Stable Immobilization Matrix for Enhanced Ethylene Production: A
Framework for Photosynthetic Solid-State Cell Factories. Green Chem.
2021, 23, 3715−3724.
(12) Lapponi, M. J.; Méndez, M. B.; Trelles, J. A.; Rivero, C. W. Cell
Immobilization Strategies for Biotransformations. Current Opinion in
Green and Sustainable Chemistry 2022, 33, No. 100565.
(13) Martinez-Garcia, F. D.; Fischer, T.; Hayn, A.; Mierke, C. T.;
Burgess, J. K.; Harmsen, M. C. A Beginner’s Guide to the
Characterization of Hydrogel Microarchitecture for Cellular Applica-
tions. Gels 2022, 8, 1−20.
(14) Caliari, S. R.; Burdick, J. A. A Practical Guide to Hydrogels for
Cell Culture. Nat. Methods 2016, 13, 405−414.
(15) Patiño Vargas, M. I.; Martinez-Garcia, F. D.; Offens, F.;
Becerra, N. Y.; Restrepo, L. M.; van der Mei, H. C.; Harmsen, M. C.;
van Kooten, T. G.; Sharma, P. K. Viscoelastic Properties of Plasma-
Agarose Hydrogels Dictate Favorable Fibroblast Responses for Skin
Tissue Engineering Applications. Biomaterials Advances 2022, 139,
No. 212967.
(16) Hitchcock, A.; Hunter, C. N.; Canniffe, D. P. Progress and
Challenges in Engineering Cyanobacteria as Chassis for Light-Driven
Biotechnology. Microbial Biotechnology 2020, 13, 363−367.
(17) Luan, G.; Zhang, S.; Lu, X. Engineering Cyanobacteria Chassis
Cells toward More Efficient Photosynthesis. Curr. Opin. Biotechnol.
2020, 62, 1−6.
(18) Fu, W.; Chaiboonchoe, A.; Khraiwesh, B.; Nelson, D. R.; Al-
Khairy, D.; Mystikou, A.; Alzahmi, A.; Salehi-Ashtiani, K. Algal Cell
Factories: Approaches, Applications, and Potentials. Marine Drugs
2016, 14, 1−19.
(19) Rosenberg, J. N.; Oyler, G. A.; Wilkinson, L.; Betenbaugh, M. J.
A Green Light for Engineered Algae: Redirecting Metabolism to Fuel
a Biotechnology Revolution. Curr. Opin. Biotechnol. 2008, 19, 430−
436.
(20) Miao, R.; Xie, H.; Liu, X.; Lindberg, P.; Lindblad, P. Current
Processes and Future Challenges of Photoautotrophic Production of
Acetyl-CoA-Derived Solar Fuels and Chemicals in Cyanobacteria.
Curr. Opin. Chem. Biol. 2020, 59, 69−76.
(21) Khan, S.; Fu, P. Biotechnological Perspectives on Algae: A
Viable Option for next Generation Biofuels. Curr. Opin. Biotechnol.
2020, 62, 146−152.
(22) Noreña-Caro, D.; Benton, M. G. Cyanobacteria as Photo-
autotrophic Biofactories of High-Value Chemicals. Journal of CO2
Utilization 2018, 28, 335−366.
(23) Guarnieri, M. T.; Pienkos, P. T. Algal Omics: Unlocking
Bioproduct Diversity in Algae Cell Factories. Photosynth. Res. 2015,
123, 255−263.
(24) Wijffels, R. H.; Kruse, O.; Hellingwerf, K. J. Potential of
Industrial Biotechnology with Cyanobacteria and Eukaryotic Micro-
algae. Curr. Opin. Biotechnol. 2013, 24, 405−413.
(25) Moreno-Garrido, I. Microalgae Immobilization: Current
Techniques and Uses. Bioresour. Technol. 2008, 99, 3949−3964.
(26) Eroglu, E.; Smith, S. M.; Raston, C. L. Application of Various
Immobilization Techniques for Algal Bioprocesses. In Biomass and
Biofuels from Microalgae; Moheimani, N. R., McHenry, M. P., de Boer,
K., Bahri, P. A., Eds.; Springer: Cham, 2015; pp. 19−44.
DOI: 10.1007/978-3-319-16640-7_2.
(27) Caldwell, G. S.; In-Na, P.; Hart, R.; Sharp, E.; Stefanova, A.;
Pickersgill, M.; Walker, M.; Unthank, M.; Perry, J.; Lee, J. G. M.
Immobilising Microalgae and Cyanobacteria as Biocomposites: New
Opportunities to Intensify Algae Biotechnology and Bioprocessing.
Energies 2021, 14, 2566.
(28) Kosourov, S.; Murukesan, G.; Seibert, M.; Allahverdiyeva, Y.
Evaluation of Light Energy to H2 Energy Conversion Efficiency in
Thin Films of Cyanobacteria and Green Alga under Photoautotrophic
Conditions. Algal Research 2017, 28, 253−263.
(29) Kosourov, S. N.; Seibert, M. Hydrogen Photoproduction by
Nutrient-Deprived Chlamydomonas Reinhardtii Cells Immobilized
within Thin Alginate Films under Aerobic and Anaerobic Conditions.
Biotechnol. Bioeng. 2009, 102, 50−58.
(30) Valsami, E. A.; Pateraki, A.; Melis, A.; Ghanotakis, D. F.
Heterologous β-Phellandrene Production by Alginate Immobilized
Synechocystis Sp. PCC 6803. J. Appl. Phycol. 2021, 33, 2157−2168.
(31) Malik, S.; Hagopian, J.; Mohite, S.; Lintong, C.; Stoffels, L.;
Giannakopoulos, S.; Beckett, R.; Leung, C.; Ruiz, J.; Cruz, M.; Parker,
B. Robotic Extrusion of Algae-Laden Hydrogels for Large-Scale
Applications. Global Challenges 2020, 4, 1900064.
(32) Wei, J.; Zhang, X.; Ai, S.; Huang, Y.; Yang, X.; Mei, Y.; Zhang,
K.; Wang, H. The Effective Astaxanthin Productivities of Immobilized
Haematococcus Pluvialis with Bacterial Cellulose. Bioresour. Technol.
2022, 344, No. 126317.
(33) Lee, H.; Shin, D.; Choi, J.; Ki, C. S.; Hyun, J. Mimicry of the
Plant Leaf with a Living Hydrogel Sheet of Cellulose Nanofibers.
Carbohydr. Polym. 2022, 290, No. 119485.
(34) Jämsä, M.; Kosourov, S.; Rissanen, V.; Hakalahti, M.; Pere, J.;
Ketoja, J. A.; Tammelin, T.; Allahverdiyeva, Y. Versatile Templates
from Cellulose Nanofibrils for Photosynthetic Microbial Biofuel
Production. J. Mater. Chem. A 2018, 6, 5825−5835.
(35) Kontturi, E.; Laaksonen, P.; Linder, M. B.; Nonappa; Gröschel,
A. H.; Rojas, O. J.; Ikkala, O. Advanced Materials through Assembly
of Nanocelluloses. Adv. Mater. 2018, 30, 1703779.
(36) Aarstad, O.; Heggset, E. B.; Pedersen, I. S.; Bjørnøy, S. H.;
Syverud, K.; Strand, B. L. Mechanical Properties of Composite
Hydrogels of Alginate and Cellulose Nanofibrils. Polymer 2017, 9,
378.
(37) Lee, K. Y.; Mooney, D. J. Alginate: Properties and Biomedical
Applications. Progress in Polymer Science (Oxford) 2012, 37, 106−126.
(38) Heise, K.; Kontturi, E.; Allahverdiyeva, Y.; Tammelin, T.;
Linder, M. B.; Nonappa; Ikkala, O. Nanocellulose: Recent
Biomacromolecules pubs.acs.org/Biomac Article
https://doi.org/10.1021/acs.biomac.3c00261
Biomacromolecules 2023, 24, 3484−3497
3496
Fundamental Advances and Emerging Biological and Biomimicking
Applications. Adv. Mater. 2021, 33, 2004349.
(39) Pawar, S. N.; Edgar, K. J. Alginate Derivatization: A Review of
Chemistry, Properties and Applications. Biomaterials 2012, 33, 3279−
3305.
(40) Wurm, F.; Rietzler, B.; Pham, T.; Bechtold, T. Multivalent Ions
as Reactive Crosslinkers for Biopolymers�A Review. Molecules 2020,
25, 1840.
(41) Hakalahti, M.; Salminen, A.; Seppälä, J.; Tammelin, T.;
Hänninen, T. Effect of Interfibrillar PVA Bridging on Water Stability
and Mechanical Properties of TEMPO/NaClO2 Oxidized Cellulosic
Nanofibril Films. Carbohydr. Polym. 2015, 126, 78−82.
(42) Lever, J.; Krzywinski, M.; Altman, N. Points of Significance:
Principal Component Analysis. Nat. Methods 2017, 14, 641−642.
(43) Altman, N.; Krzywinski, M. Points of Significance: Clustering.
Nat. Methods 2017, 14, 545−546.
(44) Luukkonen, P.; Maloney, T.; Rantanen, J.; Paulapuro, H.;
Yliruusi, J. Interaction � A Novel Approach Using Thermoporosim-
etry. Pharm. Res. 2001, 18, 1562−1569.
(45) Maloney, T. C. Thermoporosimetry of Hard (Silica) and Soft
(Cellulosic) Materials by Isothermal Step Melting. J. Therm. Anal.
Calorim. 2015, 121, 7−17.
(46) Guccini, V.; Phiri, J.; Trifol, J.; Rissanen, V.; Mousavi, S. M.;
Vapaavuori, J.; Tammelin, T.; Maloney, T.; Kontturi, E. Tuning the
Porosity, Water Interaction, and Redispersion of Nanocellulose
Hydrogels by Osmotic Dehydration. ACS Applied Polymer Materials
2022, 4, 24−28.
(47) Rippka, R.; Deruelles, J.; Waterbury, J. B. Generic Assignments,
Strain Histories and Properties of Pure Cultures of Cyanobacteria. J.
Gen. Microbiol. 1979, 111, 1−61.
(48) Saito, T.; Nishiyama, Y.; Putaux, J. L.; Vignon, M.; Isogai, A.
Homogeneous Suspensions of Individualized Microfibrils from
TEMPO-Catalyzed Oxidation of Native Cellulose. Biomacromolecules
2006, 7, 1687−1691.
(49) Scandinavian Pulp Paper and Board Testing Committee. Pulp
Total Acidic Group Content - Conductometric Titration Method; SCAN-
CM 65:02. 2002.
(50) Van De Ven, T. G. M. The Flow of Suspensions. In Polymer
Composites; The University of Wales Institute of Non-Newtonian
Fluid Mechanics, Department of Mathematics, University of Wales
Aberystwyth: Penglais, Aberystwyth, Dyfed, Wales, SY23 3BZ, 1985;
Vol. 6, pp. 209−214. DOI: 10.1002/pc.750060405.
(51) Bro, R.; Smilde, A. K. Principal Component Analysis. Anal.
Methods 2014, 6, 2812−2831.
(52) Geladi, P. Chemometrics in Spectroscopy. Part 1. Classical
Chemometrics. Spectrochimica Acta - Part B Atomic Spectroscopy 2003,
58, 767−782.
(53) Ward, J. H. Hierarchical Grouping to Optimize an Objective
Function. J. Am. Stat. Assoc. 1963, 58, 236−244.
(54) Mustila, H.; Paananen, P.; Battchikova, N.; Santana-Sánchez,
A.; Muth-Pawlak, D.; Hagemann, M.; Aro, E. M.; Allahverdiyeva, Y.
The Flavodiiron Protein Flv3 Functions as a Homo-Oligomer during
Stress Acclimation and Is Distinct from the Flv1/Flv3 Hetero-
Oligomer Specific to the O2 Photoreduction Pathway. Plant Cell
Physiol. 2016, 57, 1468−1483.
(55) Burlacot, A.; Burlacot, F.; Li-Beisson, Y.; Peltier, G. Membrane
Inlet Mass Spectrometry: A Powerful Tool for Algal Research.
Frontiers in Plant Science 2020, 11, 1−15.
(56) Beckmann, J.; Lehr, F.; Finazzi, G.; Hankamer, B.; Posten, C.;
Wobbe, L.; Kruse, O. Improvement of Light to Biomass Conversion
by De-Regulation of Light-Harvesting Protein Translation in
Chlamydomonas Reinhardtii. J. Biotechnol. 2009, 142, 70−77.
(57) Lichtenthaler, H. K. Chlorophylls and Carotenoids: Pigments
of Photosynthetic Biomembranes. In Methods in Enzymology;
Academic Press, 1987; Vol. 148, pp. 350−382. DOI: 10.1016/
0076-6879(87)48036-1.
(58) Gudin, C.; Chaumont, D. Cell Fragility - The Key Problem of
Microalgae Mass Production in Closed Photobioreactors. Bioresour.
Technol. 1991, 38, 145−151.
(59) Barbosa, M. J.; Hadiyanto; Wijffels, R. H. Overcoming Shear
Stress of Microalgae Cultures in Sparged Photobioreactors. Biotechnol.
Bioeng. 2004, 85, 78−85.
(60) Posten, C. Design Principles of Photo-Bioreactors for
Cultivation of Microalgae. Eng. Life Sci. 2009, 9, 165−177.
(61) Hubbe, M. A.; Tayeb, P.; Joyce, M.; Tyagi, P.; Kehoe, M.;
Dimic-Misic, K.; Pal, L. Rheology of Nanocellulose-Rich Aqueous
Suspensions: A Review. BioResources 2017, 12, 9556−9661.
(62) van Vliet, T. Large Deformation and Fracture Behaviour of
Gels. Curr. Opin. Colloid Interface Sci. 1996, 1, 740−745.
(63) van Vliet, T.; Luyten, H.; Walstra, P. Fracture and Yielding of
Gels. In Food Polymers, Gels and Colloids; Dickinson, E., Ed.;
Woodhead Publishing, 1991; pp. 392−403. DOI: 10.1533/
9781845698331.392.
(64) Li, M. C.; Wu, Q.; Moon, R. J.; Hubbe, M. A.; Bortner, M. J.
Rheological Aspects of Cellulose Nanomaterials: Governing Factors
and Emerging Applications. Adv. Mater. 2021, 33, 2006052.
(65) Grant, G. T.; Morris, E. R.; Rees, D. A.; Smith, P. J. C.; Thom,
D. Biological Interactions between Polysaccharides and Divalent
Cations: The Egg-Box Model. FEBS Lett. 1973, 32, 195−198.
(66) Braccini, I.; Pérez, S. Molecular Basis of Ca2+-Induced
Gelation in Alginates and Pectins: The Egg-Box Model Revisited.
Biomacromolecules 2001, 2, 1089−1096.
(67) Donati, I.; Holtan, S.; Mørch, Y. A.; Borgogna, M.; Dentini, M.;
Skjak̊-Bræk, G. New Hypothesis on the Role of Alternating Sequences
in Calcium-Alginate Gels. Biomacromolecules 2005, 6, 1031−1040.
(68) Stokke, B. T.; Draget, K. I.; Smidsrød, O.; Yuguchi, Y.;
Urakawa, H.; Kajiwara, K. Small-Angle X-Ray Scattering and
Rheological Characterization of Alginate Gels. 1. Ca-Alginate Gels.
Macromolecules 2000, 33, 1853−1863.
(69) Fang, Y.; Al-Assaf, S.; Phillips, G. O.; Nishinari, K.; Funami, T.;
Williams, P. A.; Li, A. Multiple Steps and Critical Behaviors of the
Binding of Calcium to Alginate. J. Phys. Chem. B 2007, 111, 2456−
2462.
(70) Zhang, J.; Daubert, C. R.; Foegeding, E. A. Fracture Analysis of
Alginate Gels. J. Food Sci. 2005, 70, e425−e431.
(71) Dong, H.; Snyder, J. F.; Williams, K. S.; Andzelm, J. W. Cation-
Induced Hydrogels of Cellulose Nanofibrils with Tunable Moduli.
Biomacromolecules 2013, 14, 3338−3345.
(72) Hakalahti, M.; Faustini, M.; Boissier̀e, C.; Kontturi, E.;
Tammelin, T. Interfacial Mechanisms of Water Vapor Sorption into
Cellulose Nanofibril Films as Revealed by Quantitative Models.
Biomacromolecules 2017, 18, 2951−2958.
(73) Ketola, A. E.; Leppänen, M.; Turpeinen, T.; Papponen, P.;
Strand, A.; Sundberg, A.; Arstila, K.; Retulainen, E. Cellulose
Nanofibrils Prepared by Gentle Drying Methods Reveal the Limits
of Helium Ion Microscopy Imaging. RSC Adv. 2019, 9, 15668−15677.
(74) Korhonen, J. T.; Hiekkataipale, P.; Malm, J.; Karppinen, M.;
Ikkala, O.; Ras, R. H. A. Inorganic Hollow Nanotube Aerogels by
Atomic Layer Deposition onto Native Nanocellulose Templates. ACS
Nano 2011, 5, 1967−1974.
(75) Robitzer, M.; David, L.; Rochas, C.; Di Renzo, F.; Quignard, F.
Nanostructure of Calcium Alginate Aerogels Obtained from Multistep
Solvent Exchange Route. Langmuir 2008, 24, 12547−12552.
(76) Xie, M.; Olderøy, M. O.; Zhang, Z.; Andreassen, J. P.; Strand,
B. L.; Sikorski, P. Biocomposites Prepared by Alkaline Phosphatase
Mediated Mineralization of Alginate Microbeads. RSC Adv. 2012, 2,
1457−1465.
(77) Nikkanen, L.; Santana Sánchez, A.; Ermakova, M.; Rögner, M.;
Cournac, L.; Allahverdiyeva, Y. Functional Redundancy between
Flavodiiron Proteins and NDH-1 in Synechocystis Sp. PCC 6803.
Plant Journal 2020, 103, 1460−1476.
(78) Santana-Sanchez, A.; Solymosi, D.; Mustila, H.; Bersanini, L.;
Aro, E. M.; Allahverdiyeva, Y. Flavodiiron Proteins 1−to-4 Function
in Versatile Combinations in O2 Photoreduction in Cyanobacteria.
eLife 2019, 8, 1−22.
Biomacromolecules pubs.acs.org/Biomac Article
https://doi.org/10.1021/acs.biomac.3c00261
Biomacromolecules 2023, 24, 3484−3497
3497