Cryoconcentration of Flavonoid Extract for Enhanced Biophotovoltaics and pH

Sensitive Thin Films

A. Demirbas
Agricultural and Biological Engineering, Institute of Food and Agricultural Sciences, University of Florida, Gainesville, FL

K. Groszman
Department of Computational and Applied Mathematics, Rice University, Houston, TX
~o-Hernandez
M. Pazmin
Agricultural and Biological Engineering, Institute of Food and Agricultural Sciences, University of Florida, Gainesville, FL
D. C. Vanegas
Food Engineering Department, Universidad del Valle, Cali Colombia
B. Welt
Agricultural and Biological Engineering, Institute of Food and Agricultural Sciences, University of Florida, Gainesville, FL

J. A. Hondred
Mechanical Engineering Department, Iowa State University, Iowa City, IA

N. T. Garland
Mechanical Engineering Department, Iowa State University, Iowa City, IA

J. C. Claussen
Mechanical Engineering Department, Iowa State University, Iowa City, IA
E. S. McLamore
Agricultural and Biological Engineering, Institute of Food and Agricultural Sciences, University of Florida, Gainesville, FL

DOI 10.1002/btpr.2557
Published online October 4, 2017 in Wiley Online Library (wileyonlinelibrary.com)

Flavonoids are important value added products for dye sensitized solar cells biosensors,
functional foods, medicinal supplements, nanomaterial synthesis, and other applications. Bras-
sica oleracea contains high levels of anthocyanins in leaf sap vacuoles, and there are many via-
ble extraction techniques that vary in terms of simplicity, environmental impact, cost, and
extract photochemical/electrochemical properties. The efficiency of value added biotechnolo-
gies from flavonoid is a function of anthocyanin activity/concentration and molecule stability
(i.e., ability to retain molecular resonance under a wide range of conditions). In this paper, we
show that block cryoconcentration and partial thawing of anthocyanin from B. oleracea is a
green, facile, and highly efficient technique that does not require any special equipment or pro-
tocols for producing enhanced value added products. Cryoconcentration increased anthocya-
nin activity and total phenol content approximately 10 times compared with common extraction
techniques. Cryoconcentrated extract had enhanced electrochemical properties (higher oxida-
tion potential), improved chroma, and higher UV absorbance than extract produced with other
methods for a pH range of 2–12, with minimal effect on the diffusion coefficient of the extract.
As a proof of concept for energy harvesting and sensor applications, dye sensitized solar cells
and pH-sensitive thin films were prepared and tested. These devices were comparable with
other recently published biotechnologies in terms of efficacy, but did not require expensive/
environmentally detrimental extraction or concentration methods. This low cost, biorenewable,
and simple method can be used for development of a variety of value added products. V C 2017

American Institute of Chemical Engineers Biotechnol. Prog., 34:206–217, 2018

Keywords: anthocyanin, cryoconcentration, dye sensitized solar cell, pH sensitive thin film

Introduction
Additional Supporting Information may be found in the online ver-
sion of this article.
In plants, flavins are involved in photoprotection and photo-
Correspondence concerning this article should be addressed to E. S. transduction,1 nitrogen translocation,2 and pollinator attraction.3
McLamore at emclamor@ufl.edu Flavins include carotenoids (liposoluble) and anthocyanins

206 C 2017 American Institute of Chemical Engineers

V
Biotechnol. Prog., 2018, Vol. 34, No. 1 207

(water soluble), each class having a range of antioxidant and well as total phenol content were monitored over time. Next,
photochemical properties.4,5 Plants that are rich in anthocyanins the most efficient extraction/concentration method was ana-
include blueberries, raspberries, concord grapes, and red cab- lyzed using standard electrochemical and photochemical
bage.6 Anthocyanin extracts are used as food additives,7,8 techniques. Finally, we demonstrate potential applications by
medicinal supplements,9,10 molecular templates for nanomate- developing and testing a dye sensitized solar cell and a pH
rial synthesis,11,12 various cosmetic applications,13 dye- sensitive thin film using cabbage extract.
sensitized solar cells,14,15 and as biodetectors.16,17 Due to the
diverse applications of anthocyanins, research on the develop-
ment of this value added agricultural product has increased in Methodology
the last decade.18–20 Reagents and chemicals
The general photochemistry of anthocyanins is well Red cabbage (Brassica oleracea var. capitata, f, rubra) was
known.21,22 Due to the aromatic cyclic arrangement, antho- purchased from an organic market in Gainesville, FL during
cyanins absorb low energy radiation, which varies depending the spring. Methanol (CH3OH), acetic acid (CH3COOH),citric
on pH as a function of molecular resonance. Some anthocya- acid (C6H5O327 ), disodium phosphate (Na2HPO4), potassium
nins have demonstrated fluorescence with emission near ferrocyanide (KFeCN6), lithium perchlorate (ClLiO4), sodium
600–630 nm,23,24 while some inhibit fluorescence of biocon- carbonate (Na2CO3), potassium nitrate (KNO3) potassium
jugates such as DNA–propidium iodide.25 Red cabbage chloride (KCl), sodium hydroxide (NaOH), hydrochloric acid
(Brassica oleracea L. var. capitata) can have up to 36 differ- (HCl), sodium acetate (CH3COONa), sodium bicarbonate
ent anthocyanins, each of which vary in terms of physico- (NaHCO3), sodium carbonate (Na2CO3), potassium iodide,
chemical properties and stability.26 B. oleracea anthocyanins Folin–Ciocalteu reagent, and biological buffers (MES, BES,
are primarily composed of cyanidin or peonidin based agly- TRIZMA, AMPSO, CAPS, and CABS) were purchased from
cones, although pelargonidin has been reported as well. Most Sigma-Aldrich (St. Louis, MO).DPPH (1,1-diphenyl-2-picryl-
B. oleracea anthocyanins are mono or di-acylated anthocya- hydrazyl) was purchased from Fisher Scientific (Atlanta, GA).
nins, with acid moieties consisting of sinapic, ferulic, caffeic Trolox (6-hydroxy-2,5,7,8-tetramethylchroman-2-carboxylic
and p-coumaric acids. acid) was purchased from Alexis (Axxora, Switzerland). Plati-
Anthocyanins from B. oleracea are encased in leaf sap num wire (0.3 mm diameter) was purchased from Alfa Aesar
vacuoles and can be extracted using high pressure extrac- (Ward Hill, MD), and working electrodes (Pt/Ir and glassy car-
tion,27 microwave irradiation,28 supercritical fluid extrac- bon) were purchased from Bioanalytical Systems (West Lafa-
tion,29 Soxhlet/solvent extraction,27 adsorbent purification,14 yette, IN). Titanium dioxide (TiO2) nanoparticle paste
ultrasonic treatment,30 or pulsed electric field processing,31 (Nanoxide-T, colloidal anatase with particle size of 13 nm)
among other techniques. The antioxidant and optical proper- was obtained from Solaronix (Aubonne, Switzerland).
ties of extracts vary widely for these methods, even for the
same species and harvest season.32 The efficacy of many
biotechnologies that utilize anthocyanins as a working com- Preparation of extract
ponent of the device (e.g., solar cells, sensors, functional B. oleracea leaves were pulled from the stem, the stem
foods) are dependent on the total concentration and chemico- was discarded, and leaves were chopped into 1 cm2 pieces
physical properties of the extract, where native physiological using an autoclaved knife. Any leaves that did not contain
structures are desired (pH 5 7, low salinity, absence of sol- purple pigment were discarded. Four different methods were
vents). Methods for increasing extract yield include passive used for extraction, including (i) solvent extraction, (ii)
cryoconcentration,33–35 vacuum-assisted cryoconcentration,36 microwave assisted extraction, (iii) conduction (boiling), and
centrifugal cryoconcentration,37 falling film freeze concentra- (iv) conduction with cryoconcentration. For solvent and
tion,38 and nanofiltration.39 For development of new technol- microwave extraction, aliquots of chopped leaves (100 g)
ogies such as sensors/biosolar cells, many of these were added to either 100 mL methanol at varying dilutions,
concentration methods are cost prohibitive, as they require or 100 mL DI water with 10% acetic acid, respectively. For
expensive consumables, high pressure, large centrifuges, or microwave-assisted extraction, the power was 1000 W and
other equipment. This high cost often impedes translation to treatment times of 0, 30, 45, 60, 90, 120, 240, or 480 s were
the industrial market, creating a bottleneck of biotechnolo- used (average final temperature for all samples was
gies. Among the concentration methods listed here, cryocon- 95 6 68C). The sample was placed in the center of the
centration with passive thawing and gravity separation is the microwave for all relevant experiments. For conduction-
most attractive green method, as it can be conducted at phys- based extraction, 100 g of fresh chopped leaves were added
iological pH in the absence of solvents. This is a critical fea- to 100 mL of DI water and samples were heated at different
ture for manufacturing of sensors/biosolar cells at any temperatures (25, 45, 60, 80, or 1008C) for either 30 min or
appreciable scale. Further, the technique uses commonly 1 h.
available equipment, and thus capital costs are low for For cryoconcentration, samples were transferred to a plas-
implementation. tic container, sealed, and stored at 2808C for at least 2 h
In this paper, we conduct a comparative study of facile immediately after extraction. No pretreatment was used for
anthocyanin extraction techniques including solvent extrac- the freeze concentrated extract, and no agitation was used
tion, microwave assisted extraction, and conductive heating. during the freezing process (known as “block concen-
We compare three parameters for variations of each tech- tration”). After block freezing, samples were stored at 208C
nique: (i) total anthocyanin content, (ii) antioxidant activity, and the extract at the bottom of the plastic container (a vis-
and (iii) total polyphenol content. To improve yield, block cous purple syrup) was collected after 5 min, 1, 5, and 24 h,
cryoconcentration followed by passive partial thawing was and analyzed immediately as noted (see highlight image).
used and then anthocyanin content, antioxidant activity as All aqueous extracts were vacuum filtered using a 0.45 lm
208 Biotechnol. Prog., 2018, Vol. 34, No. 1

cellulose acetate filter Whatman (Kent, WA) prior to was measured at k 5 760 nm using a spectrometer. Aliquots
analysis. of 5 mL samples contained 2% (w/w) sample, 5% (w/w)
Folin–Ciocalteu reagent, 15% (w/w) Na2CO3, and 78% (w/
w) DI water.
Total anthocyanin content
Total anthocyanin content was measured using the differ-
ential pH method by Lee et al.,40 which is a rapid and sim- Electrochemistry
ple spectrophotometric method based on the anthocyanin To determine the diffusion properties of the extract (an
structural transformation that occurs with a change in pig- estimate of molecular weight) as well as the general antho-
ment color at pH 1.0 versus pH 4.5. Two buffer systems cyanin content, electrochemical analysis of extract was car-
were used including potassium chloride buffer (0.025M, ried out based on published methods.44–46 For all analysis,
pH 5 1.0) and sodium acetate buffer (0.4M, pH 5 4.5). An cyclic voltammetry (CV) and linear sweep voltammetry
aliquot of the cabbage extract (1.0 mL) was placed in a (LSV) were carried out in buffer at 208C. Bulk pH in the
20 mL volumetric flask, diluted to volume with pH 1.0 electrochemical cell was monitored with a standard potentio-
buffer or pH 4.5 buffer, and mixed. Both solutions were metric glass pH probe (Orion 9109WL, Thermo Scientific)
incubated at room temperature for 20 min. Prepared solu- during all CV and LSV experiments. For studying the
tions were then measured with a UV/VIS spectrophotometer effect(s) of pH, extract (1 mL) was added to a 10 mL elec-
(Beckman Coulter DU-640 spectrophotometer, Beckman trochemical cell with buffers at various pKa, including: (i)
Instruments, CA, USA) at 510 and 700 nm. Anthocyanin 79.5 mM citric acid with 20.6 mM Na2HPO4 (pKa 5 3.0),
was calculated according to Eq. 1 and expressed as mg of ii) 61.5 mM citric acid with 38.6 mM Na2HPO4
cyanidin-3-glucoside (c3g) per gram of cabbage: (pKa 5 4.0), iii) 48.5 mM citric acid with 51.5 mM
Na2HPO4 (pKa 5 5.0), iv) MES (pKa 5 6.1), v) MOPS
A  MW  DF (pKa 5 7.2), vi) TRIZMA (pKa 5 8.1), vii) AMPSO
Antho5 (1)
EI (pKa 5 8.9), viii) CAPS (pKa 5 10.4), and ix) 90 mM
where Na2CO3 with 10 mM NaHCO3.
Antho 5 anthocyanin concentration (mg-c3g/L), All experiments used a three-electrode cell stand (Bioana-
lytical Systems, West Lafayette, IN) with a glassy carbon or
A 5 differential absorbance 5 (A510 – A700) at pH1.0 –
Pt/Ir working electrode, a 1 mm platinum wire as counter
(A510 – A700) at pH4.5,
electrode, and a Ag/AgCl reference electrode as previously
A510 5 absorbance at k 5 510 nm (arb. units), described.47–49All CV analyses were performed with a Pt/Ir
A700 5 absorbance at k 5 720 nm (arb. units), electrode in potassium ferrocyanide (4 mM) at a switching
MW 5 molecular weight (611 g/mol for c3g), potential of 600 mV with a five second quiet time at scan
DF 5 dilution factor, rates of 25–500 mV s21 as noted. LSV was performed with
I 5 path length (cm), and a glassy carbon electrode between 0 and 800 mV at a scan
rate of 5 mV s21, with lithium perchlorate (0.1 mM) as the
E’ 5 molar extinction coefficient for c3g (30,175 L 3
supporting electrolyte. The Randles–Sevcik theorem was
mol21 3 cm21).
used to calculate the diffusion coefficient for a given total
concentration based on Ref. (50).
Antioxidant activity
ip 52:693105  n3=2  A  D1=2  C  v1=2
Antioxidant activity was evaluated using a modified ver-
sion of the Trolox equivalent antioxidant capacity (TEAC) where
assay.41 Stock solutions were prepared by dissolving 24 mg ip 5 oxidative/reductive peak current (A),
DPPH with 100 mL methanol and then storing at 2208C n 5 number of electrons transferred in redox reaction,
when not in use. The working solution was obtained by mix-
ing 10 mL stock solution with 45 mL methanol to obtain an D 5 diffusion coefficient (cm2 s21),
absorbance of 1.1 6 0.02 U at 515 nm for a blank sample C 5 total anthocyanin concentration (M),
(methanol). Cabbage extract (100 lL) was reacted with A 5 electroactive surface area of working electrode
3900 lL of the DPPH solution for 60 min in the dark at (0.02 cm2), and
room temperature. Meanwhile, a 100 mL aliquot of Trolox V 5 scan rate (V s21).
solution with respective concentrations of 100, 200, 400,
600, 800, or 1000 mM was added to 3.9 mL DPPH working
Photochemistry
solution for generating the standard curve. Then, absorbance
was recorded at 515 nm and results were expressed in molar- Photochemistry was performed using an Ocean Optics
ity of Trolox equivalent per mass of wet extract. fiber optic UV–VIS spectrometer with fiber optic accessory
from 250 to 850 nm (Ocean Optics, Dunedin, FL). A P-400-
2-UV–VIS fiber optic cable (Ocean Optics) was used where
Total polyphenol content noted. Where noted, the fiber was positioned near the thin
Total phenolic content was determined using a modified film using a precision linear micromanipulator stage (2 lm
version of the Folin–Ciocalteu assay based on Stanković42 resolution) with custom fiber housing (Newport Corporation,
and Piljac et al.43 Gallic acid (GA) was used as the standard San Francisco CA). Extract (1 mL) was added to a cuvette
and all data was expressed as mass of GA equivalents per with 3 mL of buffer at various pKa values as described in
mass of dry defatted matter (mg-GAE/g-DM) based on GA the previous section. Samples were added to sterile cuvettes,
calibration curves. Samples and prepared standards were inverted to mix, and then absorbance was measured (relative
incubated for 30 min in a 408C water bath, and absorbance to a control sample of DI water). Where noted, thin films
Biotechnol. Prog., 2018, Vol. 34, No. 1 209

were measured using a 100 lm single mode silicone-coated For testing pH sensitive films, water samples from a local
UV–VIS fiber and a custom 3D printed film analysis setup. A lake, and soil slurry samples from various soil types were
boxcar width of 15 and an integration time of 100 ms was used prepared. Water samples from Lake Alice conservation area
for all samples. For analyzing color of each sample, the CIE at the University of Florida (Gainesville, FL) were acquired
LAB system was used, which is an international standard in November, 2016. The location includes 130 acres of pro-
defined by the Commission Internationale de l’Eclairage using a tected area, with an 82 acre open water system that has
chromaticity diagram. In this system, L 5 luminance, A 5 red inputs from stormwater runoff, interstorm discharges, irriga-
green chromaticity, and B 5 blue-yellow chromaticity. tion water, and direct rainfall. Samples were collected near
the southeast storm sewer drainage pipe at noon, and ana-
lyzed on site immediately. Soil samples were taken from
Proof of principle applications two vegetable garden sites (Gainesville, FL). One plot con-
There are a variety of applications for anthocyanin derived tained oak–saw palmetto scrub (Alaquods) that had compost
from cabbage beyond food additives. To show two specific (10% w/w) applied weekly for the past 5 years (noted as
applications, flavonoid-sensitized solar cells (i.e., Gr€atzel organic soil). The second sample site consisted of a garden
C
cells) were prepared based on the methods by Kumara that was prepared with commercial soil (Miracle GroV) that
et al.51 (see Supporting Information Figure S1 for schematic, contained a mixture of perlite (99.44%), ammonia nitrogen
energy diagram, and typical I–E curve). Manufacture of (0.035%), nitrate nitrogen (0.035%), phosphorous pentoxide–
these solar cells is described in detail by Kumara et al. P2O5 (0.07%), soluble potash–K2O (0.07%), and proprietary
Briefly, photo electrodes were prepared by coating pre- wetting agents (0.07%). For tests, a 1:2 (soil:electrolyte) soil
cleaned fluorine doped tin oxide (FTO) glass slurry was prepared by mixing 10 g of soil sample with
(50 mm 3 50 mm, 1.1 mm thick, 7 X sheet resistance) with 20 mL of 10 mM CaCl2. The soil slurry was subsequently
TiO2 nanoparticle paste using the doctor blade method (40– stirred vigorously and allowed to settle for 15 min. This
50 lm film thickness). Electrodes (1 cm 3 1 cm) were pre- measurement avoids bias by varying salt concentrations
heated to 508C for 20 min and then sintered at 4508C for among the soils, which is the official method adopted by the
30 min. Photoelectrodes were then immersed into extract for Association of Official Analytical Chemists.56 The liquid
20 min at room temperature, rinsed in ethanol, and then air extract was measured by either immersing the pH electrode
dried at room temperature for 30 min. Solar cells were in the sample vial or drop casting the slurry based on Miller
assembled by coupling the flavonoid-sensitized FTO elec- and Kissel.56 After drop casting the sample on the pH-
trode (anode) with a carbon-coated FTO glass slide (cathode) sensitive film, the fiber optic probe was positioned 1 mm
and clamping the ends. The electrolyte (0.5M potassium from the surface using the micromanipulator, and a measure-
iodide) was added by drop casting 50 lL onto each edge of ment taken.
the glass electrode interface, and cells were allowed to stabi-
lize at room temperature in the presence of 1000 W/m2 Statistics and data analysis
white light for 20 min. Based on I–E curves, the short circuit Analysis of variance (ANOVA model I) was performed to
current density (Jsc), open circuit voltage (Voc), maximum determine whether any effects in the optical properties of the
power output (Pmax), voltage at max power (Vmp), current films are statistically significant (a 5 0.05). Where relevant,
density at max power (Jmp), form factor (FF), and solar con- error bars represent the standard deviation of the arithmetic
version efficiency (g) were calculated using the methods mean for the random experimental design.
reported by Hug et al.52
Inkjet printing of the extract closely followed our similar
procedures for printing graphene inks.53–55 For preparing RESULTS AND DISCUSSION
inkjet-printed thin film biodetectors, aliquots (3 mL) of Extract anthocyanin and phenolic content
extract were filtered through a 0.45 mm. Polytetrafluoroethy- Figure 1 shows the average anthocyanin content, antioxi-
lene (PTFE) syringe filter and then loaded into an inkjet dant activity, and total phenol content for extractions using
printer cartridge (1.5 mL max. fill) and then printed through various solvent extraction (100% methanol), microwave irra-
inkjet printer (Fujifilm Dimatix DMP-2850) using a piezo- diation (8 min) and boiling methods. For variations of each
electric driven nozzle with a 10 pL nominal drop volume. method see the Supporting Information section (Supporting
The substrate holder (vacuum plate) was maintained at 508C Information Figures S2–S4). The average anthocyanin con-
to ensure uniform, rapid drying of the printed ink. Extract tent for solvent extraction at room temperature, microwave
was inkjet printed with a 20 mm drop spacing and 25 passes. irradiation, and conduction for 1 h were not significantly dif-
These printer settings permit an even printing of the extract ferent (P 5 0.15, a 5 0.05). The anthocyanin content (Fig-
without holes/gaps in the film and without material pooling ure 1a) for solvent (methanol) extraction, microwave (MW)
effects at the edges. The printed line morphology resulting (120 s), MW 1 acetic acid (10%), and boiling were not sig-
from the drop spacing and the working temperature control nificantly different. Microwave irradiation for 8 min or boil-
edge effects (known as the coffee ring effect). Our previous ing for 60 min increased anthocyanin content by at least
work shows that average film thickness is between 3 and 35% compared with the other methods (MW for 8 min and
5 mm, classifying this structure as a thin film.53–55 A circular boiling for 1 h were not statistically different (P 5 0.001,
array was designed using computer-aided design (CAD soft- a 5 0.05). Antioxidant activity (Figure 1b) and total phenol
ware) and inkjet printed. For unbuffered calibration tests, pH content (Figure 1c) were at least 40% higher for 60 min of
was adjusted with 3M HCl or 2M NaOH, and then 5 mL ali- boiling compared with solvent extraction, although this was
quots were drop cast onto the pH sensitive film. For buffered not statistically different than boiling for 30 min, MW
calibration tests, 10 mM stock solutions of buffer (as (8 min) 1 AA (10%), or MW (8 min). Among all extraction
described previously) and drop cast as described above. methods, the highest absorption (k 5 435 nm) was for
210 Biotechnol. Prog., 2018, Vol. 34, No. 1

Figure 2. Photochemical behavior for various extraction meth-

ods.
(a) UV VIS spectra (kEx 5 280 nm) for cryoconcentrated
Figure 1. Comparison of the best extraction methods used in extract at various pH. (b) Average peak intensity at 420, 450,
this study for anthocyanin content (a), antioxidant and 640 nm for cryoconcentrated extract (kEx 5 280 nm). (c)
activity (b), and total phenol content (c). Other Representative UV VIS spectra for boiled extract, microwave
with acetic acid (MW 1 AA), and cryoconcentrated extract
methods can be found in the supplemental section.
(kEx 5 300 nm); photograph shows color of extract after
Different lower case letters denote statistically differ- extraction. (d) Average peak intensity for various extraction
ent groups (ANOVA, a 5 0.05). methods shown in c for an excitation of 300 nm. The absorp-
tion intensity for cryoconcentrated extract was significantly
extract that was boiled for 60 min (1420 a.u.; CIE 5 550). higher than all other methods (P  0.01, a 5 0.05).
Although temperatures of 1008C can indeed denature extract
proteins57 and anthocyanin,58 boiling produced the optimal non-neutral pH (methanol extraction) or using microwave
result when considering anthocyanin content, antioxidant irradiation (with or without AA) resulted in lower phenol
activity, total phenol concentration, absorption, and total and anthocyanin relative to boiling for 60 min (for photos of
cost. Preliminary analysis with dye-sensitized solar cells extracts see inset in Figure 2c). Although the specific molec-
(DSSC) and pH-sensitive thin films showed that extraction at ular mechanism for this reduction in DSSC/sensor efficiency
Biotechnol. Prog., 2018, Vol. 34, No. 1 211

Figure 3. Effect of thawing time on cryoconcentrated extract, including: anthocyanin content (a), antioxidant activity (b), and total
phenol content (c). The highest concentration of each was recorded after 5 min of thawing, which decreased at an average
rate of 0.6 6 0.2 h21. Different lower case letters denote statistically different groups (ANOVA, a 5 0.05).

is not known, this result is most likely due to the presence We speculate that the marked difference in dilution by water is
of compounds not analyzed by the assays used in this study the main reason for the increased phenol content and anthocya-
as the CIE data indicate clear differences in the pigmentation nin activity in the extract, although the specific mechanism is
of each extract. Other established techniques such as adsorp- not known at this time. Aider and de Halleux33 showed that,
tion column separation could be used to improve the purifi- other than the total time to thaw, there was no difference
cation of the extract, removing impurities such as sugars, between microwave-assisted thawing or passive thawing of
organic acids, and proteins,14 but these techniques can be cryoconcentrated maple sap extract (the study measured sugar
cost prohibitive in many applied settings, such as develop- content, electroconductivity, and CIE chromaticity). However,
ment of low cost sensors. in our work we saw a clear difference in the extracts taken at
To improve the extraction yield, block cryoconcentration different times during passive block thawing.
followed by passive partial thawing was used (Figure 3). The results in Figures 1–3 indicate that boiling for one hour
Miyawaki59 showed that passive (i.e., gravitational) partial followed by cryoconcentration enhances extraction efficiency
thawing of the cryoconcentrated block is optimal for high sol- considerably. Our study did not use vacuum-assisted cryocon-
ute concentrations, while final freezing temperature had no centration or tubular ice systems for progressive cryoconcen-
effect on extract yield. Anthocyanin content/antioxidant activ- tration due to the added expense and need for additional
ity and total phenol content were measured over time during equipment, although both methods may possibly improve
passive thawing. After 5 min at 208C, the anthocyanin content extraction efficiency. Petzold et al.37 showed that the time for
(2649 6 83 mg/mL), antioxidant activity (1635 6 cryoconcentration can be reduced by up to 33% if a vacuum
106 lmol g21) and total phenol content (1732 6 14 mg/g) of system is used. Gunathilake et al.60 and Miyawaki et al.61
the extract were significantly higher than all other methods showed that tubular ice systems for partial ice melting enhance
tested (P < 0.01, a 5 0.05). As expected, these values extraction efficiency by up to 20% relative to progressive cryo-
decreased exponentially as the frozen extract was diluted by concentration and passive thawing. The technique here is a
melting ice. Anthocyanin content and antioxidant activity simple, highly efficient green method for obtaining highly con-
(0.6 6 0.2 and 0.2 6 0.1 h, respectively) decreased at a lower centrated anthocyanin at physiological pH and salinity.
rate than the total phenol (1.0 6 0.3 h). Surprisingly, after 24 h
at 208C, the extract was fully thawed and the anthocyanin con-
tent/activity was higher than extract prepared by boiling in Extract electrochemistry
water for 1 h (P 5 0.010, a 5 0.05), although the total phenol Figure 4a shows a representative linear sweep voltammo-
content was not statistically different (P 5 0.13, a 5 0.05). gram (LSV) at various extract concentrations (pH 5 7). The
212 Biotechnol. Prog., 2018, Vol. 34, No. 1

Figure 4. Electrochemical behavior of extract with and without cryoconcentration (sample taken 5 min after placing frozen extract at
208C).
(a) A representative linear sweep voltammogram (LSV) at various concentrations of cryoconcentrated extract (pH 5 7). Inset shows peak current at
various anthocyanin concentrations. (b) Average concentration of oxidizeable species calculated from LSV data. (c) Representative cyclic voltammo-
gram (CV) at various scan rates (pH 7). Inset shows Cottrell plot indicating a linear increase in peak oxidation current with the square root of the
scan rate. (d) Effect of pH on diffusion coefficient with and without cryoconcentration.

inset shows the relationship between average oxidative peak structure. The inset in Figure 4c shows a linear Cottrell plot,
current and total concentration of oxidizable species, which indicating that transport was diffusion limited, as expected.
increases linearly as expected. As the pH increased, the oxi- Using the Randles–Sevcik theorem, the diffusion coefficient
dation potential decreased (Figure 4b), indicating that flavo- (D) was calculated at various pH levels; data shown for cry-
noids are easily oxidized at higher pH as hydroxyl groups oconcentrated extract and boiled extract are shown in Figure
are deprotonated in accordance with the Nernst law 4d. For both extracts, the value of D slightly decreased as
(slope 5 58 6 3 V/pH). At all pH levels tested, the cryocon- the pH increased, which was also shown by Xavier et al.62
centrated extract had a significantly higher concentration of The net change in diffusion coefficient from pH 3 to pH 7
oxidizeable species than boiled extract (P < 0.0001, (12.9 6 2.1%) was more significant than the net change
a 5 0.05), which is similar to the trend in Figure 1. In addi- from pH 7 to pH 11 (5.7 6 1.6%). This is most likely attrib-
tion, the extract was more stable at lower pH, based on the uted to nonacylated anthocyanins (comprising 20–30% of
results shown here. The oxidation potential at 1 V is associ- total anthocyanins) and methylated anthocyanins (e.g., peoni-
ated with electrochemical oxidation of flavonoids (FL- din), which are sensitive to pH changes. There was no sig-
OH ! FL-O• 1 e– 1 H1); where FL is a flavonoid.45 The nificant difference between the two extraction procedures
electrochemical data in Figure 4a,b and antioxidant activity with regards to diffusion coefficient at all pH levels tested
measured by radical scavenging (DPPH) from Figure 1b (P < 0.0001, a 5 0.05). This electrochemical characteriza-
(FL-OH ! FL-O• 1 H1) are tightly correlated since each tion confirms that the cryoconcentrated extract has potential
technique reports data based on the same hydroxyl groups application as a dye in electrochemical devices.
on the anthocyanin.
Figure 4c shows the oxidative sweep of a representative
Extract photochemistry
cyclic voltammogram (CV) at various scan rates for an
extract concentration of 1.7 lM and pH 5 7. The oxidation Aqueous extracts of anthocyanin are known to undergo a
peak near 1 V confirms that functional hydroxyl groups on variety of resonant structural transformations when the pH of
anthocyanin were electrochemically oxidized at the electrode the solution changes, which is the major driving force for
surface. Based on the work by Lima et al.,45 the relatively observed color changes.63 At low pH (3–4), the anthocyanin
high oxidation potential corresponds to C-3 hydroxyl groups, was protonated (i.e., flavylium salt) and appeared red in
resornicol groups, and the presence of sugars in the ring color with a CIE chroma of 496. See Supporting Information
Biotechnol. Prog., 2018, Vol. 34, No. 1 213

Figure S5 for photographs and CIE chroma of extract at var-

ious pH levels. The peak absorption for pH 3–4 was at
420 nm for an excitation of 250 nm, followed by a broad
peak between 600 and 700 nm (Figure 2a,b). As pH was
increased to the 4–5 range, a slow color transition occurred
as the anthocyanin adopted a carbinol structure; some antho-
cyanins are known to be colorless in this state while others
are pink with a CIE chroma of 528.63 The peaks at 420, 450,
and 640 nm all decreased during this slow carbinol transi-
tion. Slight bathochromic shifts occurred as pH increased,
although this change was not significant. At neutral pH, the
anthocyanin was rapidly deprotonated and in the anhydro
base conformation (appearing bluish purple in color with a
CIE chroma of 555). The peaks at 420 and 540 nm
decreased considerably, while the peak at 450nm did not
change significantly compared with the flavylium and carbi-
nol structures. When the pH was increased to 10, a slow
transition to the chalcone conformation was obtained as the
ring structure opened, and the solution appeared green with
a CIE chroma of 485. Above pH 10 the color of the solution
was dark yellow with a chroma of 572, the peaks at 420 and
640 nm increased (similar to carbinol peaks), while the peak
at 450 nm did not change. As described by Gomes et al.,21
kinetics of these structural transitions play a major role in
the photochemical behavior. For the fast deprotonation con-
formational transitions at pH 3,6,7, and 8 there was a linear
relationship between pH and UV absorption for the 420 nm
peak (R2 5 0.90), 450 nm peak (R2 5 0.82), and 640 nm
peak (R250.92). However, for the slow transition to carbinol
and chalcone resonant structures, there was no linear rela-
tionship between peak intensity and pH. Figure 2d shows
that the average peak absorbance for the cryoconcentrated
extract was significantly higher than other methods
(P < 0.0001, a 5 0.05), indicating that the extract has excel-
lent potential for colorimetric biosensing.

Proof of principle applications

As discussed in detail by Hug et al.,52 DSSC have been
developed with a variety of anthocyanins, including extract
from B. oleracea, black rice (Oryza sativa), Manchu rose
(Rosa xanthina), spiderwort (Tradescantia zebrine), and Rho- Figure 5. Demonstration of DSSC based on cryoconcentration
dodendron, among others. Figure 5a shows the electrochemi- when compared with heat extraction (boiling).
cal behavior (I–E curve) and power output of a DSSC at (a) Current–potential (I–E) curve and power curve for a
flavonoid-sensitized solar cell. (b) I–E and power curve for cry-
various illumination levels with and without cryoconcentra- oconcentrated extract at various pH levels, showing an increase
tion. The short circuit current density (JSC), open circuit in power output with decreasing pH. (c) Schematic of energy
potential (EOC), maximum power output (Pmax), and solar diagram for DSSC depicting electron transfer between the
conversion efficiency (g) increased with increasing illumina- anode and cathode using iodide as the electrolyte.
tion, as expected (see Supporting Information Tables S1–S2
for details). The form factor (FF) for all DSSC was not sig-
nificantly different, but all other solar power conversion performance, which is similar to the work by Hemmatzadeh
parameters were significantly higher for cryoconcentrated et al.65 comparing various solvent extraction techniques for
extract compared with boiled extract (P 5 0.001, a 5 0.05). DSSC (including ethanol extraction and heat extraction). In
These results are similar to other DSSC reviewed by Hug addition to the effect of anthocyanin content on DSSC per-
et al.,52 most of which used pretreatment methods, expensive formance, pH also plays an important role in DSSC. Figure
materials, or complex electrodes. On the other hand, our 5b shows that efficiency increased by 20% when the pH was
device was developed by simply concentrating the anthocya- decreased from 7 to 3 due to electron injection at the metal
nin and using simple assembly methods as described in oxide surface using boronic acid as a dye anchor to the
detail by Furukawa et al.64 The highest efficiency was mea- metal oxides.66
sured for DSSC prepared with cryoconcentrated extract at an Although we did not test the effect of immersion time on
illumination of 200 mW (49%), which was over four times performance (anodes were immersed for 30 min in this
higher than the efficiency of DSSC prepared with boiled study), Li et al.67 showed that DSSC power increases with
extract. This shows a clear relationship between anthocyanin time as the anode is immersed in extract for longer periods,
content, antioxidant activity (recall Figure 1) and DSSC with maximum power measured after 24 h, implying a
214 Biotechnol. Prog., 2018, Vol. 34, No. 1

Figure 6. Inkjet-printed red cabbage extract on nanocellulose substrate in a circular array for pH sensing.
(a) Photograph of inkjet-printed “dots” on a circular disc. The outer ring represents pH increasing clockwise from 1 to 14, the middle ring represents
control samples (pH 7), the third ring represents increase of pH clockwise from 7 to 14, and the middle circle is pH 1. (b) UV–VIS spectra for pH 2
to pH 6; (c) VIS spectra for pH 6–11. The pH-sensitive film was applied for analyzing water samples from (d) a fresh water lake, (e) a commercial
soil, and (f) an organic soil. CIE and peak absorption at 435 nm are shown for each thin film. Reference image (buffers) are shown for comparison.

concentration dependence. Since the DSSC efficiency ring). The middle ring of dots shows a control (pH 5 7), the
depends on the concentration of flavonoids, the cryoconcen- inner ring shows standards of pH 14–7, and the innermost
tration method is a highly useful and simple mechanism for dot is pH 1. The color of the films in Figure 6 are nearly
enhancing DSSC performance. In the future, DSSC can be identical to the trend in Figure 2; red at pH < 4, blue at pH
further improved by deposition of nanocrystalline TiO2 5–6, purple at pH 7, blue/green at pH 8–9, and green/yellow
films,68 use of plastic substrates such as polyethylene naptha- above pH 10. In Figure 6b,c, the UV–VIS spectra are shown
late,69 or by mixing extracts from multiple plant species (white light excitation) for pH 2–6. The violet peak
such as red cabbage and blueberry.70 Figure 5c shows the (435 nm) and cyan peak (480 nm) also followed the same
energy pathway for the electrochemical cell, and a detailed trends in Figure 2, although the bathochromic shifts were
schematic of the working mechanism can be seen in Sup- more pronounced than when the anthocyanin was suspended
porting Information Figure S1. These results clearly show in aqueous solution. Figure 6d–f shows applications of the
that cryoconcentration of extract enhances DSSC perfor- printed anthocyanin using samples from lake water (Figure
mance when compared with boiled extract. 5d), commercial soil (Figure 6e), and organic soil (Figure
Figure 6 shows the performance of inkjet-printed biodetec- 6f). In each case, the reference image from Figure 6a, the
tors for direct measurement of pH. Figure 6a shows a photo- measured CIE, and the absorbance (kEx 5 435 nm) are
graph of the inkjet-printed “dots” at pH from 1 to 14 (outer shown for comparison. For each of the samples, the pH
Biotechnol. Prog., 2018, Vol. 34, No. 1 215

measured with a standard electrode was within 6 0.5 pH Literature Cited

units (see Supporting Information Table S3 for details). 1. Briggs WR, Liscum E. The role of mutants in the search for the
These results demonstrate that the inkjet-printed films are photoreceptor for phototropism in higher plants. Plant Cell
applicable for rapid field monitoring of solution pH in sam- Environ. 1997;20:768–772.
ples relevant to agricultural, ecological, and environmental 2. Masclaux-Daubresse C, Chardon F. Exploring nitrogen remobili-
studies. zation for seed filling using natural variation in Arabidopsis
thaliana. J Exp Bot. 2011;62:2131–2142.
In addition to the proof of principle demonstrations herein, 3. Gonzalez-Teuber M, Heil M. Nectar chemistry is tailored for
our previous work has shown that anthocyanin can be used both attraction of mutualists and protection from exploiters.
to synthesize nanoparticles12 or for functional foods,9 and Plant Signal Behav 2009;4:809–813.
Casta~neda-Ovando et al.4 discuss a variety of other applica- 4. Castaneda-Ovandode Pacheco-Hernnndez Lourdes AM, Paez-
Hernandez ME, Rodriguez JA, Galnn-Vidal Andres C. Chemical
tions such as cosmetic and pharmaceutical uses. We have studies of anthocyanins: a review. Food Chem. 2009;113:859–
shown the simple extraction and cryoconcentration of antho- 871.
cyanin from B. oleracea using block cryoconcentration with 5. Dang K-V, Plet J, Tolleter D, Jokel M, Cuine S, Carrier P,
partial gravitational thawing as an effective method for creat- Auroy P, Richaud P, Johnson X, Alric J, Allahverdiyeva Y,
ing value added products. In addition to producing highly Peltier G. Combined increases in mitochondrial cooperation and
oxygen photoreduction compensate for deficiency in cyclic elec-
stable extract with attractive properties, the leaves are intact
tron flow in Chlamydomonas reinhardtii. Plant Cell 2014;26:1–
after boiling, and could be consumed or used as a food prod- 16.
uct, when compared with solvent extraction methods which 6. Fossen T, Cabrita L, Andersen OM. Colour and stability of pure
destroy the food product. Other applications not discussed in anthocyanins influenced by pH including the alkaline region.
detail here include use of cabbage extract as a biosorbent71 Food Chem. 1998;63:435–440.
and cancer preventative medicines.72 The low cost, simple 7. Dyrby M, Westergaard N, Stapelfeldt H. Light and heat sensitiv-
ity of red cabbage extract in soft drink model systems. Food
method can be created in any lab without the use of spe- Chem. 2001;72:431–437.
cialty equipment. 8. Walkowiak-Tomczak D, Czapski J. Colour changes of a prepa-
ration from red cabbage during storage in a model system. Food
Chem. 2007;104:709–714.
Conclusion 9. Wang J, Huang Y, Li K, Chen Y, Vanegas DC, McLamore ES,
Shen Y. Leaf extracts from Lithocarpus polystachyus Rehd. pro-
There is a need for sustainable value added products and mote glycogen synthesis in T2DM mice. PLoS One.
technologies derived from agricultural materials. Among the DOI:10.1371/journal.pone.0166557, 2016.
10. Nabavi SF, Habtemariam S, Daglia M, Shafighi N, Barber AJ,
diverse library of bioderived materials from agricultural feed-
Nabavi SM. Anthocyanins as a potential therapy for diabetic ret-
stocks, flavonoids, including anthocyanins, are among one of inopathy. Curr Med Chem. 2015;22:51–58.
the most promising due to low cost and land use at the farm 11. Xie Y, Kocaefe D, Chen C, Kocaefe Y. Review of research on
scale. In this work, we show that cryoconcentration of leaf template methods in preparation of nanomaterials. J Nanomater.
extract from red cabbage (B. oleracea) is a facile, low cost, 2016;2016:1–10.
green approach for producing large quantities of anthocyanin. 12. Demirbas A, Welt BA, Ocsoy I. Biosynthesis of red cabbage
extract directed Ag NPs and their effect on the loss of antioxi-
The extract has excellent photochemical and electrochemical dant activity. Mater Lett. 2016;179:20–23.
stability, and can be used in a number of applications. To dem- 13. AnangaGeorgiev A, Ochieng V, Phills JB, Tsolov V. Production
onstrate this concept, two proof of principle technologies were of anthocyanins in grape cell cultures: a potential source of raw
fabricated and tested (a dye sensitized solar cell and a thin film material for pharmaceutical, food, and cosmetic industries. Med-
biodetector). These results clearly show that cryoconcentration iterr Genet Code Grapevine Olive. 2013;247–287.
14. Jampani C, Raghavarao KSMS. Process integration for purifica-
of B. oleracea leaf extract is a useful method that can be used tion and concentration of red cabbage (Brassica oleracea L.)
to develop biotechnologies such as solar cells and sensors. anthocyanins. Sep Purif Technol. 2015;141:10–16.
Future research will investigate the stability of cabbage extract 15. Zhu H, Zeng H, Subramanian V, Masarapu C, Hung K-H, Wei
from cryoconcentration compared to other methods for creat- B. Anthocyanin-sensitized solar cells using carbon nanotube
ing value added products. films as counter electrodes. Nanotechnology 2008;19:465204.
16. Siedler S, Stahlhut SG, Malla S, Maury J, Neves AR. Novel
biosensors based on flavonoid-responsive transcriptional regula-
tors introduced into Escherichia coli. Metab Eng. 2014;21:2–8.
Acknowledgments 17. Dhale R, Dhongade V, Aiyer RC, Kodam K. Anthocyanin
The authors acknowledge the UF Student Science Training coated carbon dioxide biosensor. 2nd Int Symp Phys Technol
Sens. 2015;8–10.
Program (KG), the Colombian funding agencies Colfuturo y 18. Diaz JT, Chinn MS, Den Truong V. Simultaneous saccharifica-
Colciencias (DV-No.001375882), the Agricultural and Bio- tion and fermentation of industrial sweetpotatoes for ethanol
logical Engineering Department at the University of Florida production and anthocyanins extraction. Indus Crops Prod.
(ESM-USDA/NIFA Hatch Project No. FLA-ABE-005062), 2014;62:53–60.
the Turkish Ministry of National Education (AD),the 19. Kautsky H, Hirsch A. Energie-Umwandlungen an Grenzfl€achen,
IV. Ber Dtsch Chem Ges B 64:2677–2683.
National Institute of Food and Agriculture, US Department
20. Chen H, Yada R. Nanotechnologies in agriculture: new tools for
of Agriculture, under award number 11901762 (JCC); the sustainable development. Trends Food Sci Technol. 2011;22:
Iowa State University College of Engineering and Depart- 585–594.
ment of Mechanical Engineering(JCC), and the USDA Mul- 21. Gomes R, Parola AJ, Lima JC, Pina F. Solvent effects on the
tistate Project (NC1194) for financial support. The authors thermal and photochemical reactions of 4-Iodo-8-
also thank Y. Yagiz from the Food and Environmental Toxi- methoxyflavylium and their consequences on the coloring phe-
nomena caused by anthocyanins in plants. Chem A Eur J. 2006;
cology Lab (FETL) in the Department of Food Science and 12:7906–7912.
Human Nutrition at the University of Florida for assistance 22. Quina FH, Moreira PF, Vautier-Giongo C, Rettori D, Rodrigues
with antioxidant testing. RF, Freitas AA, Silva PF, Maçanita AL. Photochemistry of
216 Biotechnol. Prog., 2018, Vol. 34, No. 1

anthocyanins and their biological role in plant tissues. Pure 44. Aguirre MJ, Chen YY, Isaacs M, Matsuhiro B, Mendoza L,
Appl Chem 2009;81:1687–1694. Torres S. Electrochemical behaviour and antioxidant capacity of
23. Drabent R, Pliszka B, Olszewska T. Fluorescence properties of anthocyanins from Chilean red wine, grape and raspberry. Food
plant anthocyanin pigments. I. Fluorescence of anthocyanins in Chem. 2010;121:44–48.
Brassica oleracea L. extracts. J Photochem Photobiol B Biol. 45. de Lima A, Sussuchi EM, Giovani WFD. Electrochemical and
1999;50:53–58. antioxidant properties of anthocyanins and anthocyanidins.
24. Roshchina VV, Roshchina VV. Vital autofluorescence: applica- Croat Chem Acta 2007;80:29–34.
tion to the study of plant living cells. Int J Spectrosc. 2012; 46. Moncada MC, de Mesquita MF, dos Santos MMC. Electrochem-
2012:1–14. ical oxidation of the synthetic anthocyanin analogue 4-methyl-
25. Bennett MD, Price HJ, Johnston JS. Anthocyanin inhibits propi- 7,8-dihydroxyflavylium salt. J Electroanal Chem. 2009;636:60–
dium iodide DNA fluorescence in Euphorbia pulcherrima: 67.
implications for genome size variation and flow cytometry. Ann 47. Vanegas DC, Taguchi M, Chaturvedi P, Burrs S, Tan M,
Bot. 2008;101:777–790. Yamaguchi H, McLamore ES. A comparative study of carbon–
26. Charron CS, Clevidence BA, Britz SJ, Novotny JA. Effect of platinum hybrid nanostructure architecture for amperometric
dose size on bioavailability of acylated and nonacylated antho- biosensing. Analyst 2014;139:660–667.
cyanins from red cabbage (Brassica oleracea L. Var. capitata). 48. Chaturvedi P, Vanegas DC, Taguchi M, Burrs SL, Sharma P,
J Agric Food Chem. 2007;55:5354–5362. McLamore ES. A nanoceria-platinum-graphene nanocomposite
27. Arapitsas P, Turner C. Pressurized solvent extraction and mono- for electrochemical biosensing. Biosens Bioelectron. 2014;58:
lithic column-HPLC/DAD analysis of anthocyanins in red cab- 179–185.
bage. Talanta 2008;74:1218–1223. 49. Burrs SL, Vanegas DC, Bhargava M, Mechulan N, Hendershot
28. Lu LZ, Zhou YZ, Zhang YQ, Ma YL, Zhou LX, Li L, Zhou P, Yamaguchi H, Gomes C, McLamore ES. A comparative
ZZ, He TZ. Anthocyanin extracts from purple sweet potato by study of graphene-hydrogel hybrid bionanocomposites for bio-
means of microwave baking and acidified electrolysed water sensing. Analyst 2015;140:1466–1476.
and their antioxidation in vitro. Int J Food Sci Technol. 2010; 50. McLamore ES, Shi J, Jaroch D, Claussen JC, Uchida a, Jiang
45:1378–1385. Y, Zhang W, Donkin SS, Banks MK, Buhman KK, Teegarden
29. Khoddami A, Wilkes MA, Roberts TH. Techniques for analysis D, Rickus JL, Porterfield DM. A self referencing platinum nano-
of plant phenolic compounds. Molecules 2013;18:2328–2375. particle decorated enzyme-based microbiosensor for real time
30. €
Demird€oven A, Ozdo gan K, Erdogan-Tokatlı K. Extraction of measurement of physiological glucose transport. Biosens Bioe-
anthocyanins from red cabbage by ultrasonic and conventional lectron. 2011;26:2237–2245.
methods: optimization and evaluation. J Food Biochem. 2015; 51. Kumar S, McEvoy N, Lutz T, Keeley GP, Nicolosi V, Murray
39:491–500. CP, Blau WJ, Duesberg GS. Gas phase controlled deposition of
31. Gachovska T, Cassada D, Subbiah J, Hanna M, Thippareddi H, high quality large-area graphene films. Chem Commun. (Camb)
Snow D. Enhanced anthocyanin extraction from red cabbage 2010;46:1422–1424.
using pulsed electric field processing. J Food Sci. 2010;75:323– 52. Hug H, Bader M, Mair P, Glatzel T. Biophotovoltaics: natural
329. pigments in dye-sensitized solar cells. Appl Energy 2014;115:
32. Park S, Arasu MV, Jiang N, Choi SH, Lim YP, Park JT, Al- 216–225.
Dhabi NA, Kim SJ. Metabolite profiling of phenolics, anthocya- 53. He Q, Das SR, Garland NT, Jing D, Hondred JA, Cargill AA,
nins and flavonols in cabbage (Brassica oleracea var. capitata). Ding S, Karunakaran C, Claussen JC. Enabling inkjet printed
Indus Crops Prod. 2014;60:8–14. graphene for ion selective electrodes with post-print thermal
33. Aider M, de Halleux D. Passive and microwave-assisted thaw- annealing. ACS Appl Mater Interfaces. DOI:10.1021/acsa-
ing in maple sap cryoconcentration technology. J Food Eng. mi.7b0009, 2017.
2008;85:65–72. 54. Das SR, Uz M, Ding S, Lentner MT, Hondred JA, Cargill AA,
34. Aider M, de Halleux D, Melnikova I. Gravitational and Sakaguchi DS, Mallapragada S, Claussen JC. Electrical differen-
microwave-assisted thawing during milk whey cryoconcentra- tiation of mesenchymal stem cells into Schwann-cell-like pheno-
tion. J Food Eng. 2008;88:373–380. types using inkjet-printed graphene circuits. Adv Heal Mater.
35. Aider M, de Halleux D. Production of concentrated cherry and DOI: 10.1002/adhm.201601087.
apricot juices by cryoconcentration technology. LWT Food Sci 55. Das SR, Nian Q, Cargill AA, Hondred JA, Ding S, Saei M,
Technol. 2008;41:1768–1775. Cheng GJ, Claussen JC. 3D nanostructured inkjet printed gra-
36. Petzold G, Orellana P, Moreno J, Cerda E, Parra P. Vacuum- phene via UV-pulsed laser irradiation enables paper-based elec-
assisted block freeze concentration applied to wine. Innov Food tronics and electrochemical devices. Nanoscale 2016;8:15870–
Sci Emerg Technol. 2016;36:330–335. 15879.
37. Petzold G, Aguilera JM. Centrifugal freeze concentration. Innov 56. Miller RO, Kissel DE. Comparison of soil pH methods on soils
Food Sci Emerg Technol. 2013;20:253–258. of North America. Soil Sci Soc Am J. 2010;74:310.
38. Gulfo R, Auleda JM, Moreno FL, Ruiz Y, Hernandez E, 57. Aider M, de Halleux D, Akbache A. Whey cryoconcentration
Raventos M. Multi-plate freeze concentration: recovery of sol- and impact on its composition. J Food Eng. 2007;82:92–102.
utes occluded in the ice and determination of thawing time. 58. Arkoub-DjermouneBoulekbache-Makhlouf L, Zeghichi-Hamri L,
Food Sci Technol Int. 2014;20:405–419. Bellili S, Boukhalfa SF, Madani K. Influence of the thermal
39. Cisse Vaillant MF, Pallet D, Dornier M. Selecting ultrafiltration processing on the physico-chemical properties and the antioxi-
and nanofiltration membranes to concentrate anthocyanins from dant activity of a solanaceae vegetable: eggplant. J Food Qual
Roselle extract (Hibiscus sabdariffa L.). Food Res Int. 2011;44: 2016.
2607–2614. 59. Miyawaki O, Kato S, Watabe K. Yield improvement in progres-
40. Lee J, Durst RW, Wrolstad RE. Determination of total mono- sive freeze-concentration by partial melting of ice. J Food Eng.
meric anthocyanin pigment content of fruit juices, beverages, 2012;108:377–382.
natural colorants, and wines by the pH differential method: col- 60. Gunathilake M, Dozen M, Shimmura K, Miyawaki O. An appa-
laborative study. J AOAC Int. 2005;88:1269–1278. ratus for partial ice-melting to improve yield in progressive
41. Brand-Williams W, Cuvelier ME, Berset C. Use of a free radi- freeze-concentration. J Food Eng. 2014;142:64–69.
cal method to evaluate antioxidant activity. LWT Food Sci Tech- 61. Miyawaki O, Gunathilake M, Omote C, Koyanagi T, Sasaki T,
nol. 1995;28:25–30. Take H, Matsuda A, Ishisaki K, Miwa S, Kitano S. Progressive
42. Milan SS. Total phenolic content, flavonoid concentration and freeze-concentration of apple juice and its application to pro-
antioxidant activity of Marrubium peregrinum L. extracts. Kra- duce a new type apple wine. J Food Eng. 2016;171:153–158.
gujev J Sci. 2011;33:63–72. 62. Xavier MF, Lopes TJ, Quadri MGN, Quadri MB. Extraction of
43. 
Piljac-Zegarac J, Stipčević2 T, Belsčak A. Antioxidant proper- red cabbage anthocyanins: optimization of the operation condi-
ties and phenolic content of different floral origin honeys. tions of the column process. Braz Arch Biol Technol. 2008;51:
J ApiProduct ApiMed Sci. 2009;1:43–50. 143–152.
Biotechnol. Prog., 2018, Vol. 34, No. 1 217

63. Olivas-Aguirre FJ, Rodrigo-Garcia J, Martinez-Ruiz NDR, as photosensitizers for dye-sensitized solar cells. J Sol-Gel Sci
Cardenas-Robles AI, Mendoza-Diaz SO, Alvarez-Parrilla E, Technol. 2013;66:212–219.
Gonzlez-Aguilar GA, De La Rosa LA, Ramos-Jimeinez A, 69. Yoo K, Kim JY, Lee JA, Kim JS, Lee DK, Kim K, Kim JY,
Wall-Medrano A. Cyanidin-3-O-glucoside: physical-chemistry, Kim B, Kim H, Kim WM, Kim JH, Ko MJ. Completely trans-
foodomics and health effects. Molecules 2016;21. parent conducting oxide-free and flexible dye-sensitized solar
64. Furukawa S, Iino H, Iwamoto T, Kukita K, Yamauchi S. Char- cells fabricated on plastic substrates. ACS Nano.2015;9:3760–
acteristics of dye-sensitized solar cells using natural dye. Thin 3771.
Solid Films 2009;518:526–529. no. p 70. Syafinar R, Gomesh N, Irwanto M, Fareq M, Irwan YM. Poten-
65. Hemmatzadeh R, Jamali A. Enhancing the optical absorption of tial of purple cabbage, coffee, blueberry and turmeric as nature
anthocyanins for dye-sensitized solar cells: enhancing the optical based dyes for dye sensitized solar cell (DSSC). Energy Proc.
absorption of anthocyanins for dye-sensitized solar cells. 2015;79:799–807.
71. Hossain M. a, Ngo HH, Guo WS, Nguyen TV, Vigneswaran S.
J Renew Sustain Energy 2015;13120.
Performance of cabbage and cauliflower wastes for heavy metals
66. Zhang L, Cole JM. Anchoring groups for dye-sensitized solar
removal. Desalin Water Treat. 2013:1–17.
cells. ACS Appl Mater Interfaces 2015;7:3427–3455. 72. Beecher CWW. Cancer preventive properties of varieties of
67. Li M, Liu Y, Wang H, Shen H. Synthesis of TiO2 submicro- Brassica oleracea: a review. Am J Clin Nutr. 1994;59.
rings and their application in dye-sensitized solar cell. Appl
Energy 2011;88:825–830.
68. Gokilamani N, Muthukumarasamy N, Thambidurai M, Ranjitha Manuscript received Mar. 25, 2017, and revision received Aug. 10,
A, Velauthapillai D. Utilization of natural anthocyanin pigments 2017.