Carbohydrate Polymers 233 (2020) 115859

Contents lists available at ScienceDirect

Carbohydrate Polymers
journal homepage: www.elsevier.com/locate/carbpol

A naked-eye detection polyvinyl alcohol/cellulose-based pH sensor for T

intelligent packaging
Lei Dinga,b, Xiang Lia,b, Lecheng Hua,b, Yunchong Zhanga,b, Yang Jianga,b, Zhiping Maoa,b,
Hong Xua,b, Bijia Wanga,b,*, Xueling Fenga,b, Xiaofeng Suia,b,*
a
Key Lab of Science and Technology of Eco-textile, Ministry of Education, College of Chemistry, Chemical Engineering and Biotechnology, Donghua University, Shanghai,
201620, People’s Republic of China
b
Innovation Center for Textile Science and Technology of DHU, Donghua University, Shanghai, 201620, People’s Republic of China

A R T I C LE I N FO A B S T R A C T

Keywords: Naked-eye detection pH sensor is becoming a powerful tool in food safety monitoring. In this work, a pH sensor
Naked-eye detection pH sensor was developed by incorporating cellulose modified with acidochromic dye into polyvinyl alcohol (PVA). The
Acidochromic cellulose results indicated that the dye (up to 170.4 μmol/g) was successfully anchored to cellulose. It was demonstrated
Leakage-resistance that the addition of acidochromic regenerated cellulose (ARC) resulted in enhancement of tensile strength,
Food monitoring
elongation at break and maximum decomposition temperature by 44 %, 43.6 % and 11 °C, respectively. The pH
sensor demonstrated that a visible color change from yellow to brick-red and to purple when placed in solutions
of pH = 7, 10 and 12. The pH sensor showed excellent resistance to leaching under strong acidic and alkaline
conditions. When applied to spoiled shrimp, an evident color change from yellow to brown was observed,
suggesting it could serve as an easy-to-use, non-destructive visual indicator system for real-time food monitoring.

1. Introduction Some researchers have constructed visual pH sensors for food safety
monitoring now. Ma, Du, and Wang (2017) prepared a biosensor by
At present, food safety issues caused by spoilage are becoming in- mixing pH responsive curcumin with tara gum/polyvinyl alcohol for
creasingly severe and it is imperative to strengthen safety monitoring in the detection of shrimp spoilage. Pereira, de Arruda, and Stefani (2015)
the process of food transportation, preservation and use (Jia et al., embedded pH sensitive anthocyanins in chitosan/polyvinyl alcohol to
2019; Kuswandi, Jayus Restyana, Abdullah, Heng, & Ahmad, 2012). construct time-temperature indicators for monitoring milk spoilage.
Gas chromatography, liquid chromatography, electrochemistry, and Nopwinyuwong, Trevanich, and Suppakul (2010) studied dessert spoi-
fluorescent probes have been used to study the freshness of food lage by a colorimetric pH sensor based on bromothymol blue, methyl
(Fiddes, Chang, & Yan, 2014; Önal, Tekkeli, & Önal, 2013; Zhang, Hou, red and polyethylene nylon. However, these sensors obtained by di-
Xie, & Gao, 2019). However these methods are complex, inconvenient rectly blending pH-responsive substance with matrix usually suffer
and require specific instruments. Spoiled food (such as fish, shrimp, from leakage problems. Therefore, it is necessary to develop a simple,
crab and meat) releases organic amines which could lead to pH change sensitive, leakage-free, low-cost and visual pH sensor.
of the surrounding environment. Thus, it is feasible and easy to study Cellulose, as the most abundant renewable resource, is in-
food spoilage by monitoring pH change (Cao et al., 2019; Liang, Sun, exhaustible, degradable, biocompatible and easy to modify (Ding et al.,
Cao, Li, & Wang, 2018; Wei, Cheng, Ho, Tsai, & Mi, 2017). Visual pH 2018; Rol, Belgacem, Gandini, & Bras, 2019; Wang, Lu, & Zhang, 2016).
sensors are smart materials that signal pH change by changing color. In the past decades, some efforts have been devoted to developing vi-
They have a unique advantage for being non-destructive and enable sual cellulosic pH sensors by anchoring the acidochromic substance to
naked-eye detection without using any instruments (Pacquit et al., cellulose (Cao et al., 2019; Chauhan et al., 2014; Devarayan & Kim,
2007; Zhang, Lu, & Chen, 2014). In recent years, visual pH sensors have 2015; Kassal et al., 2017). These sensors were resistant to leakage and
received much attention in fields ranging from food safety, environ- inherited the favorable characteristics of cellulose and the pH re-
mental monitoring, analytical chemistry, and biomedicine (Devadhasan sponsors. However, their sensitivity to water and flexibility need to be
& Kim, 2015; Silva-Pereira, Teixeira, Pereira-Júnior, & Stefani, 2015). improved before they can find applications as intelligent packaging

⁎
Corresponding authors at: Key Lab of Science and Technology of Eco-textile, Ministry of Education, College of Chemistry, Chemical Engineering and
Biotechnology, Donghua University, Shanghai, 201620, People’s Republic of China.
E-mail addresses: [email protected] (B. Wang), [email protected] (X. Sui).

https://doi.org/10.1016/j.carbpol.2020.115859
Received 6 July 2019; Received in revised form 7 January 2020; Accepted 10 January 2020
Available online 13 January 2020
0144-8617/ © 2020 Elsevier Ltd. All rights reserved.
L. Ding, et al. Carbohydrate Polymers 233 (2020) 115859

Fig. 1. Schematic illustration of the preparation of naked-eye detection PVA/ARC pH film sensor.

materials (Bardet, Belgacem, & Bras, 2015; Chauhan et al., 2014; Liu 2.2. Preparation of RC suspension
et al., 2009; Liu, Guo, Nan, Duan, & Zhang, 2017; Shi et al., 2011; Xiong
et al., 2014). This problem could be solved by blending the acid- The cotton fabrics (2 g) were grounded in a MF 10 grinder (IKA,
ochromic cellulose with a suitable polymer. PVA is commonly used Germany) for 3 min (1000 W, 3000 rpm). The resulting powder was
synthetic polymer that is biocompatible, biodegradable, nontoxicity, stirred in 18 mL D.I. water for 30 min in a three-necked flask equipped
and has good processability and mechanical properties. It also has ex- with a mechanical stirrer. Precooled phosphoric acid (0 °C, 300 mL, 85
cellent miscibility with biopolymers including cellulose, chitosan and %) was then added and stirring was continued at 4 °C for 24 h to fully
starch, and is widely applied in the fields of packaging materials dissolve cellulose. The cellulose solution was poured into water
(Guzman-Puyol et al., 2015; Jiang et al., 2018; Ma, Du, Yang et al., (1500 mL) to form a milky suspension. The phosphoric acid in the
2017; Youssef & El-Sayed, 2018). suspension was removed by repeated centrifugation (12,000 rpm,
In this work, a naked-eye detection pH sensor was developed by Thermo, USA) until the suspension was neutral. Finally, the RC sus-
combining PVA and regenerated cellulose (RC) grafted with an acid- pension was passed through a high pressure homogenizer (APV-2000,
ochromic dye (Fig. 1). ARC was fully characterized by FT-IR, XPS, 13C Germany) set at 500 bar for 3 times.
NMR, TEM, XRD, and thermal analysis, and its color change properties
were studied. A strong and flexible pH responsive film sensor was then
2.3. Preparation of ARC suspension
obtained by blending ARC and PVA. This sensor had very low leakage
under strong acid and alkaline conditions and was used to monitor
The acidochromic dye 1-hydroxy-4-[4-(ethylsulfurate sulfonyl)-
shrimp spoilage. It shows great potential in real-time monitoring of
phenylazo]-naphthalene (HESPN) was synthesized by a method similar
food spoilage.
to that reported in the literature (Mohr, Werner, & Wolfbeis, 1994). ESA
(5 g) was added to HCl solution (25 %, 6 g), and the mixture was cooled
to 4 °C and stirred for 1 h in ice bath. NaNO2 (1.375 g) was dissolved in
2. Materials and methods
water (5 g), and added dropwise to the ESA/HCl solution. The reaction
mixture was stirred for 1.5 h in ice bath to obtain the diazo component.
2.1. Materials
1-Naphthol (2.875 g) and acetic acid (50 mL) as coupling component
were added to a three-necked flask, to which the diazo component was
The plain woven cotton fabrics (107 g/m2) were obtained from Hua
added dropwise. The mixture was stirred for 4 h at 20 °C, and the re-
Fang Co., Ltd (Binzhou, China). Phosphoric acid (85 %) and 1-naphthol
sulting acidochromic HESPN dye was filtered out, washed with acetic
were purchased from Shanghai Tian Scientific Co., Ltd (Shanghai,
acid, and dried in air. 1H-NMR (400 MHz, DMSO-d6, ppm): 8.60 (d, 1H,
China). 4-(ethylsulfurate sulfonyl) aniline (ESA) was supplied by Hubei
=CH-), 7.59–8.24 (m, 8H, =CH-), 6.84 (d, 1H, =CH-), 5.48 (s, 1H,
Chushengwei Chemical Co., Ltd (Hubei, China). Sodium nitrite
-SO3H), 3.45–3.71 (m, 4H, −CH2CH2-).
(NaNO2), hydrochloric acid (37 %, HCl), sodium hydroxide (NaOH),
RC suspension (2.5 wt%, 40 g), NaCl (4.80 g) and acidochromic
sodium chloride (NaCl), sodium carbonate (Na2CO3) and PVA
HESPN dye (0, 20, 50 and 100 mg) were combined in a reactor and hold
(1750 ± 50) were supplied by Sinopharm Chemical Reagent Co., Ltd
at 30 °C for 10 min. Na2CO3 (1.20 g) was added to the mixtures which
(Shanghai, China). Dimethyl sulfoxide (DMSO), acetic acid and am-
were heated at a rate of 2 °C/min from 30 to 65 °C. Then the reaction
monia solution (25 %) were purchased from Shanghai Lingfeng
was continued at 65 °C for 90 min. The mixtures were subsequently
Chemical Reagent Co., Ltd (Shanghai, China). All chemicals were used
cooled to room temperature and the products were sequentially washed
as received.
with water, DMSO, water, and finally passed through a high pressure
homogenizer (APV-2000, Germany) set at 500 bar for 3 times. The re-
sulting products were coded as RC, ARC-1, ARC-2 and ARC-3,

2
L. Ding, et al. Carbohydrate Polymers 233 (2020) 115859

respectively. was drawn. The dye fixation content, dye fixation rate and degree of
substitution were calculated according to the absorption of the solution
2.4. Preparation of PVA/ARC pH film sensors and the standard curve of the HESPN dye.

PVA solution (3 wt%, 20 g) was mixed with ARC-3 suspension 2.5.8. Thermogravimetric analysis (TGA)
(0.8 wt%, 0 g, 1.875 g, 3.750 g and 7.500 g) at room temperature, and Thermal behaviors of the RC, ARC, PVA and PVA/ARC samples
stirred at 200 rpm for 30 min. The mixture were then degassed by de- were conducted by a thermogravimetric analyzer (TG 209F1,
compression method for 10 min to obtain homogeneous dispersions. In Germany). The experiments were performed under nitrogen atmo-
order to get the pH film sensors, the dispersion was cast onto mould and sphere from 30 to 600 °C at 10 °C /min. The derivative of thermal
dried at 50 °C for 24 h (humidity 15 %). The resulting film products analysis curves of the samples were also recorded.
were coded as PVA, PVA/ARC-2.5 %, PVA/ARC-5 % and PVA/ARC-10
%, respectively. 2.5.9. Mechanical properties
The tensile strength and elongation at break of PVA and PVA/ARC
2.5. Characterizations films were evaluated using a universal testing machine (UH6502,
China) with a strain rate of 10 mm/min. Before testing, the films were
2.5.1. Fourier transform infrared spectroscopy (FT-IR) and Nuclear cut into 100 mm × 10 mm and placed for 24 h in a constant tempera-
magnetic resonance spectroscopy (NMR) ture and humidity (20 °C, 65 % RH).
FT-IR spectra of the HESPN dye, RC and ARC were recorded from
500 to 4000 cm−1 using a PerkingElmer Spectrum-Two (USA) equipped 2.5.10. Leakage-resistance property
with an attenuated total reflectance (ATR) accessory. The dry RC and In order to evaluate the leakage-resistance property of HESPN dye
ARC samples were prepared by freeze-drying their suspensions. immobilized in pH film sensor, the PVA/ARC-5 % sample (0.1 g) was
The 1H NMR and solid state 13C NMR spectra were recorded on a immersed in acid-base solutions (10 mL, pH = 2 and 12). The solutions
Bruker AVANCE 400 spectrometer (Switzerland). were measured by a UV–vis spectrophotometer (PerkinElmer, USA) at
time intervals of 1, 3 and 5 days.
2.5.2. X-ray photoelectron spectroscopy (XPS)
XPS spectra of the RC and ARC were obtained with an X-ray pho- 2.5.11. Detection of ammonia solution
toelectron spectrometer (Thermo ESCALAB 250, USA) equipped with The PVA/ARC films were immersed in different concentrations of
monochromatic Al Kα source (1486.6 eV). ammonia solution for 5 min, and then removed to observe their color
and test their UV–vis absorption spectra.
2.5.3. Transmission electron microscopy (TEM)
Morphology of the RC and ARC was studied using transmission 2.5.12. Monitoring of shrimp spoilage
electron microscope (JEM-2100, Japan). The samples were diluted to Shrimps (25 g) and PVA/ARC pH film sensor (1 cm × 1 cm) were
0.1 mg/mL, then dropped and dried on a carbon-coated copper grid placed on a plate for 24 h at 25 °C. The color parameters L (Lightness), a
(200 mesh). (Red-Green) and b (Yellow-Blue) of the sensor before and after the
spoilage test were recorded using a spectrophotometer (Datacolor 650,
2.5.4. X-ray diffraction (XRD) America).
XRD patterns of the RC and ARC were recorded using an X-ray
diffractometer (Rigaku D/max 2550PC, Japan) equipped with Cu Kα 2.5.13. Statistical analysis
radiation (40 kV and 200 mA) in a range of 2θ from 5° to 60°. According The samples were tested at least three times, and the results were
to the XRD pattern, crystallinity index (CI) of sample was calculated by reported as mean ± SD. The statistical differences were analyzed by
the following equation (Eq. 1), the one-way ANOVA, and the significance was defined at P < 0.05.
IMax − IAm
CI = ×100% 3. Results and discussion
IMax (1)
13
Where IMax represented the maximum intensity of crystalline cellulose, 3.1. FT-IR, XPS and C NMR analyses of ARC
and IAm represented the intensity of amorphous cellulose.
The optical image of a fully washed ARC initially showed the
2.5.5. Zeta potential HESPN dye on the ARC after dyeing, as shown in Fig. S1. Then, the FT-
The zeta potential of RC and ARC dispersion (0.1 wt%) were mea- IR spectra of the HESPN dye, RC and ARC samples are recorded and
sured by a Malvern Zetasizer instrument (Nano ZS, UK) at 25 °C. shown in Fig. 2(a). Comparing with RC, a new peak at 1546 cm−1
corresponded to vibrations of the aromatic CeC skeleton was observed
2.5.6. Contact angle (CA) in the spectra of the dye and ARC, confirming the presence of the pH
The contact angle of RC, ARC, PVA and PVA/ARC films were re- responsive dye on the RC. XPS analysis was performed to study the
corded on a DSA30 contact angle meter (Germany). The RC and ARC changes of cellulose before and after dyeing. As shown in Fig. 2(b), (c)
films were obtained by filtering their suspensions and drying. and (d), compared with RC, the presence of S2p and N1s peaks in the
ARC spectrum and the increase of the peak accounting of the aromatic
2.5.7. UV–vis analysis and Determination of grafted dye carbon at 284.8 eV in the ARC spectrum were the clear indications of
UV-vis absorption spectra of the HESPN dye, ARC and PVA/ARC pH responsive dye on the ARC. The solid-state 13C NMR spectra of RC
samples at different pH values were recorded using a UV–vis spectro- and ARC were used to confirm the successful grafting of the HESPN dye
photometer (PerkinElmer, USA). on RC, as shown in Fig. 3. Similar to the previously reported literatures,
The amount of grafted dye was measured using a previously re- the chemical shifts of carbon in RC also changed after dye modification
ported procedure (Ding et al., 2018). Specifically, the solution of pre- (Gorgieva, Vogrinčič, & Kokol, 2015; Wang, Fu, Leu, & Dong, 2018).
dried ARC powder (0.1 g) dissolved in sulfuric acid (5 mL, 70 wt%) was Compared with RC, the chemical shift of the amorphous region
diluted to a suitable volume, and then pH of the solution was adjusted (61.20 ppm) of C6 in ARC was obvious, increasing by 0.90 ppm, which
to 12 for measuring its absorption with a UV–vis spectrophotometer. was caused by the resonance up-shift effect of adjacent carbon during
The standard curve of the HESPN dye in aqueous solution (pH = 12) the hydroxyl modification. In addition, C2, C3, C5 in the ARC also

3
L. Ding, et al. Carbohydrate Polymers 233 (2020) 115859

Fig. 2. FT-IR spectra (a) of ARC, RC and HESPN dye, Survey spectra (b) and C1s high resolution region (c, d) of RC and ARC.

showed increased chemical shifts of about 0.4 ppm. These results in- were calculated and summarized in Table 1. Dye fixation content was
dicated that the HESPN dye was grafted on the RC and the dyeing re- determined from the standard curve of pH responsive HESPN dye
action was most likely to occur at the hydroxyl groups of the amor- (0.003-0.010 g/L) measured at pH = 12. The results showed the dye
phous region of C6. fixation content significantly increased from 43.7 to 170.4 μmol/g with
Dye fixation content, degree of substitution and dye fixation rate the increase of dye dosage (P < 0.05). Comparing to a previous report

13
Fig. 3. Solid-state C NMR spectra (a) and deconvolution curves (b, c) of RC and ARC.

4
L. Ding, et al. Carbohydrate Polymers 233 (2020) 115859

Table 1 and 29 %, respectively, and decreased with the increase of dye dosage.
Dye fixation content, degree of substitution and dye fixation rate for ARC. This could be a result of destruction of the hydrogen bond between
Sample Dye fixation content degree of substitution Dye fixation rate cellulose chains upon incorporation of the dye molecules.
(μmol/g) (%) The thermal stability of RC, ARC-1, ARC-2 and ARC-3 were studied
by TGA and DTG, and the data are shown in Fig. 4(d). According to the
ARC-1 43.7 ± 1.1 c 0.0071 ± 0.0002 c
74.2 ± 1.9 a
TGA and DTG curves, the maximum decomposition temperatures (Tmax)
ARC-2 103.7 ± 1.5 b 0.0168 ± 0.0003 b
70.3 ± 1.0 b

ARC-3 170.4 ± 1.4 a 0.0276 ± 0.0002 a

57.8 ± 0.5 c of the samples were all within the range of 342−348 °C, which in-
dicated dye modification barely affected the thermal stability of RC.
Different letters in the same column indicate significant differences (P < 0.05). Similar to the pure cellulose, the dye-modified ARC is also highly hy-
drophilic (Fig. S3).
using cellulose nanocrystals (Chauhan et al., 2014), the amount of dye
fixed on cellulose was significantly higher (170.4 μmol/g V.S.
3.3. pH responsive properties of ARC
19.5 μmol/g). This could be due to the lower crystallinity of RC. The
degree of substitution of cellulose was calculated and obviously in-
As shown in Fig. 5(a), the dye-modified ARC exhibited a distinctive
creased from 0.0071 to 0.0276 with the increasing initial dye dosage
color change from yellow to purple when pH was varied between 7 and
(P < 0.05), and the color of dye-modified cellulose became darker
12. To further investigate the response of ARC to pH change, UV–vis
(Fig. S2). Dye fixation rate is a key parameter for cellulose modification,
absorption spectra of ARC suspensions were recorded at different pH
since it reflects dye utilization. The fixation rates ranged from 57 %–75
values and the results were consistent with the color change in Fig. 5(a).
%, which showed that the dye utilization could remain at a relatively
As shown in Fig. 5(b), the maximum absorption wavelength (λmax)
high level.
shifted from 465 to 540 nm when the pH increased from 7 to 11, due to
the deporonation of hydroxyls in the HESPN dye. A similar shift of λmax
3.2. TEM, XRD and thermal analyses of ARC from 455 to 546 nm was observed for ARC, when the pH increased from
7 to 12 (Fig. 5(c)). The pKa value of the pH responsive dye was 8.89,
Morphological features of RC and ARC were characterized using and its UV–vis absorption was basically unchanged beyond pH = 11
TEM. As shown in Fig. 4 (a) and (b), the chemical modification on the (Fig. 5(b)). The pKa value of ARC was 10.20, the upper limit of its color
RC with pH responsive dye didn’t change the morphology of RC. Similar change range also had a slight shift to pH = 12 (Fig. 5(d)). The change
to RC, the ARC exhibited abundant networked structure, consisted of in pKa of the ARC hydroxyl could be attributed to the stabilization of
cellulose fabrils of nanometers in width and hundreds of nanometers in the hydroxyl protons via intramolecular H-bonding (Mohr et al., 1994).
length. The reversibility of ARC was studied by measuring its absorbance in
The XRD patterns of RC before and after modification with dye are HCl (pH = 5) and NaOH (pH = 12) solutions. As shown in Fig. 5(e), the
shown in Fig. 4(c). All cellulose samples displayed the same char- absorbance at 455 nm (pH = 5) and 546 nm (pH = 12) did not ob-
acteristic diffraction peaks located at 12.5° and 20.7° 2θ. These char- viously change after 10 cycles. The results clearly indicated that the
acteristic peaks were attributed to the typical crystallographic lattice of ARC had good reversible transformation between pH = 5 and pH = 12.
cellulose Ⅱ, indicating that dye modification did not change the crystal As shown in Fig. 5(f), the surface charge of ARC was affected by pH
structure of RC. According to the XRD curves (Fig. 4(c)), the calculated value and degree of substitution, and ARC-3 had the minimum zeta
crystallinity of RC, ARC-1, ARC-2 and ARC-3 were 37 %, 33 %, 32 % potential value of −19 mV at pH = 12.

Fig. 4. TEM images of RC (a) and ARC (b) suspensions; XRD (c), TGA (d) and DTG (inset of d) curves of RC, ARC-1, ARC-2 and ARC-3.

5
L. Ding, et al. Carbohydrate Polymers 233 (2020) 115859

Fig. 5. Optical images of ARC-3 suspension at different pH values (a), UV–vis spectra of pH responsive dye (b) and ARC-3 (c) at different pH values, sigmoidal fittings
of absorbance versus pH for determination of the pKa of pH responsive dye (inset of b) and ARC-3 (d), the absorbance of ARC-3 at pH = 5 (455 nm) and 12 (546 nm)
for different cycles (e), the zeta potential of RC and ARC (f). Different letters in the (e) and (f) figure indicate significant differences (P < 0.05).

3.4. Mechanical properties of pH film sensor 3.5. Thermal properties of pH film sensor

PVA was mixed with acidochromic cellulose to obtain an intelligent Thermal stability of the PVA/ARC film sensor was studied by TGA
high performance sensor. Fig. 6(a) and (b) show the tensile strength and and DTG, and compared to that of pure PVA. As shown in Fig. 6(c) and
elongation at break of pure PVA and PVA/ARC film sensors. It could be (d), three degradation stages were observed for PVA/ARC and PVA
seen from Fig. 6 that the film sensor containing 5 % ARC had the best films. Below 150 °C, there was a small weight loss of approximately 2 %
mechanical performance. Compared with pure PVA, the tensile strength due to evaporation of water. All samples lost about 68 % of their total
and elongation at break were significantly improved by 44 % and 43.6 weights between 200−350 °C mainly due to the dehydration, chain
% (P < 0.05), respectively. However, when ARC content was increased scission and decomposition of PVA (Cano, Cháfer, Chiralt, & González-
to 10 %, the tensile strength and elongation at break decreased to Martínez, 2016). When the temperature continued to rise from 350 to
33 MPa and 141 %, respectively. Tensile strength and elongation at 500 °C, PVA and PVA/ARC films had a weight loss of about 20 % due to
break are affected by dispersion and volume fraction of ARC in the the degradation of the by-products generated from PVA (Bonilla,
matrix (Niu et al., 2018). At low embedding, ARC had a good dispersion Fortunati, Atarés, Chiralt, & Kenny, 2014). From the thermal gravi-
and compatibility in the matrix allowing efficient transfer of stress from metric curves of PVA and PVA/ARC films (Fig. 6(d)), the Tmax of the
the matrix to the ARC and endowed the composite with enhanced composite films were slightly higher than that of pure PVA. And PVA/
tensile strength and flexibility. However, ARC had a strong tendency to ARC-10 % which contained the highest amount of ARC had the highest
aggregate at higher dosages of 10 % or above, and resulted in poor Tmax of 268 °C. It could be concluded that the incorporation of ARC
mechanical properties of the composite. Generally, pure cellulose films didn’t negatively affect the thermal stability of PVA.
have low elongation at break which are not conducive as packaging
materials. The PVA/ARC-5 % pH film sensor had high elongation at
3.6. pH responsive properties of pH film sensor
break (313 %) and tensile strength (36 MPa) and could meet up with
requirements for packaging materials.
The pH responsive properties of the PVA/ARC film sensors were
investigated in the range from pH = 7 to pH = 12 (Fig. 7(a) and (b)).
Similar to ARC, this film sensor exhibited yellow, brick-red and purple

6
L. Ding, et al. Carbohydrate Polymers 233 (2020) 115859

Fig. 6. Tensile strength (a), elongation at break (b), TGA (c) and DTG (d) curves of PVA, PVA/ARC-2.5 %, PVA/ARC-5 % and PVA/ARC-10 % films. Different letters
in the (a) and (b) figures indicate significant differences (P < 0.05).

at pH = 7, 10 and 12, respectively. And the UV–vis absorption spectra dye leached into the solution at Day 5, was calculated to be 0.19 ppm or
of film sensor showed that its λmax shifted from 456 nm at pH = 7 to 0.7 %, which is very low, compared to previous reported values of si-
548 nm at pH = 12. As seen in Fig. S4 and S5, the PVA/ARC film had a milar sensors (Cao et al., 2019).These results suggested that the pre-
fast response and changed color in 1 s due to its good hydrophilicity pared pH sensor film exhibited excellent resistance to leaching in the
(Fig. S6), and this favorable pH sensitivity is unvaried to film thickness strong alkaline and acid conditions.
or degree of swelling.
Leakage-resistance property of the PVA/ARC pH film sensor was
evaluated by studying the leaching of dye in acid-base solutions 3.7. Shrimp spoilage monitoring
(pH = 2 and 12), and the results are shown in Fig. 7(c). When the
sensor was immersed in the acid solution (pH = 2) for 5 days, the The PVA/ARC composite films as visual pH sensors developed in
UV–vis spectrum of immersed solution didn’t show any characteristic this study are capable of signaling in response to pH changes in the
peaks of the dye, which indicated that the sensor was not susceptible to environment. Similar to previous report (Ma, Du, Wang, 2017), here we
acid leaching. When the sensor film was immersed in the basic solution used the smart PVA/ARC film to detect the spoilage of shrimp caused by
(pH = 12), the absorbance at 533 nm gradually increased from 0.003 to the microbial activities. The formation of ammonia (NH3) and other
0.009 with prolonged immersing time from 1 to 5 days. The amount of volatile amines are the signal of the shrimp spoilage, so the response
characteristics of PVA/ARC film to NH3 concentration were

Fig. 7. Optical images (a) and UV–vis spectra (b) of PVA/ARC-5 % film sensor at different pH values, UV–vis spectra of the immersed solutions (c) and Optical images
(inset of c) of PVA/ARC-5 % film sensor immersed in acid-base solutions (pH = 2 and 12) for 1, 3 and 5 days.

7
L. Ding, et al. Carbohydrate Polymers 233 (2020) 115859

Fig. 8. UV–vis spectra (a) and optical images (b) of PVA/ARC film sensor at different concentrations of ammonia solution; Optical images of shrimp before spoilage
(c), after spoilage (d) and PVA/ARC films (e).

Table 2 Acknowledgements
The color parameters of PVA/ARC pH film sensor.
Sample L a b
This work was financially supported by the Fundamental Research
Funds for the Central Universities (No. 2232019A3-06, and No.
Initial 61.71 ± 0.61 17.77 ± 0.51 35.83 ± 0.52 2232019G-04), the Fundamental Research Funds for the Central
After spoilage 46.85 ± 1.02 * 25.02 ± 0.62 * 16.68 ± 1.40 * Universities and Graduate Student Innovation Fund of Donghua
University (No. CUSF-DH-D-2019061), and the Program of Introducing
* In the same column indicates significant differences (P < 0.05).
Talents of Discipline to Universities (No. 105-07-005735).
investigated. The PVA/ARC film was immersed in the solutions con-
Appendix A. Supplementary data
taining 0–1 × 104 ppm ammonia. The results were shown in Fig. 8(a)
and (b), which showed the PVA/ARC film was sensitive enough to
Supplementary material related to this article can be found, in the
detect ammonia at 100 ppm or above.
online version, at doi:https://doi.org/10.1016/j.carbpol.2020.115859.
As shown in Fig. 8(c) and (d), the pH sensor was used in the shrimp
detection and it changed color from yellow to brown after the sample
References
shrimps were stored for 24 h at room temperature. The color difference
of pH sensor before and after the spoilage test was summarized in
Bardet, R., Belgacem, N., & Bras, J. (2015). Flexibility and color monitoring of cellulose
Fig. 8(e) and Table 2. The L, a and b values of the pH sensor changed nanocrystal iridescent solid films using anionic or neutral polymers. ACS Applied
significantly from 61.71, 17.77 and 35.83 to 46.85, 25.02 and 16.68 Materials & Interfaces, 7(7), 4010–4018.
(P < 0.05), respectively. The results indicated that the pH sensor be- Bonilla, J., Fortunati, E., Atarés, L., Chiralt, A., & Kenny, J. M. (2014). Physical, structural
and antimicrobial properties of poly vinyl alcohol–chitosan biodegradable films. Food
came duller, redder and less yellow, and agreed well with the observed Hydrocolloids, 35, 463–470.
optical images. Therefore, the prepared PVA/ARC film could be used in Cano, A. I., Cháfer, M., Chiralt, A., & González-Martínez, C. (2016). Biodegradation be-
intelligent packaging to indicate the spoilage of shrimp and other havior of starch-PVA films as affected by the incorporation of different anti-
microbials. Polymer Degradation and Stability, 132, 11–20.
amine-releasing food via direct naked-eye inspection.
Cao, L., Sun, G., Zhang, C., Liu, W., Li, J., & Wang, L. (2019). An intelligent film based on
cassia gum containing bromothymol blue-anchored cellulose fibers for real-time
detection of meat freshness. Journal of Agricultural and Food Chemistry, 67(7),
4. Conclusions 2066–2074.
Chauhan, P., Hadad, C., Lopez, A. H., Silvestrini, S., La Parola, V., Frison, E., & Carofiglio,
T. (2014). A nanocellulose-dye conjugate for multi-format optical pH-sensing.
An intelligent PVA/ARC pH sensor for food monitoring was ob- Chemical Communications, 50(67), 9493–9496.
tained by blending PVA with RC modified with acidochromic dye. The Devadhasan, J. P., & Kim, S. (2015). An ultrasensitive method of real time pH monitoring
FT-IR, XPS and 13C NMR analyses demonstrated that the acidochromic with complementary metal oxide semiconductor image sensor. Analytica Chimica
Acta, 858, 55–59.
dye has been successfully anchored to RC, and the dye fixation content
Devarayan, K., & Kim, B. S. (2015). Reversible and universal pH sensing cellulose na-
was up to 170.4 μmol/g. The dye modification didn’t cause significant nofibers for health monitor. Sensors and Actuators B, Chemical, 209, 281–286.
change in morphology, crystallinity and thermal stability of RC. While Ding, L., Jiang, Y., Wang, B., Li, Y., Mao, Z., Xu, H., ... Sui, X. (2018). A waterborne bio-
the dye-modified ARC was sensitive to pH and exhibited a significant based polymer pigment: Colored regenerated cellulose suspension from waste cotton
fabrics. Cellulose, 25(12), 7369–7379.
color change from yellow to purple and excellent reversibility when pH Fiddes, L. K., Chang, J., & Yan, N. (2014). Electrochemical detection of biogenic amines
was varied between 5 and 12. The tensile strength, elongation at break during food spoilage using an integrated sensing RFID tag. Sensors and Actuators B,
and Tmax of the PVA/ARC film sensor were significantly improved from Chemical, 202, 1298–1304.
Gorgieva, S., Vogrinčič, R., & Kokol, V. (2015). Polydispersity and assembling phenomena
from 25 MPa, 218 % and 257 °C to 36 MPa, 313 % and 268 °C after the of native and reactive dye-labelled nanocellulose. Cellulose, 22(6), 3541–3558.
addition of ARC, respectively. Similar to ARC, the PVA/ARC film sensor Guzman-Puyol, S., Ceseracciu, L., Heredia-Guerrero, J. A., Anyfantis, G. C., Cingolani, R.,
also displayed distinct color change at different pH. This pH film sensor Athanassiou, A., & Bayer, I. S. (2015). Effect of trifluoroacetic acid on the properties
of polyvinyl alcohol and polyvinyl alcohol–cellulose composites. Chemical Engineering
had excellent leakage-resistance property, and the dye was leakage-free Journal, 277, 242–251.
under the acid solution and only 0.7 % of dye was released under the Jia, R., Tian, W., Bai, H., Zhang, J., Wang, S., & Zhang, J. (2019). Amine-responsive
alkaline solution. The PVA/ARC film sensor was used to monitor the cellulose-based ratiometric fluorescent materials for real-time and visual detection of
shrimp and crab freshness. Nature Communications, 10(1), 795.
deterioration of shrimp and its color was varied from yellow to brown Jiang, X., Xiang, N., Zhang, H., Sun, Y., Lin, Z., & Hou, L. (2018). Preparation and
after 24 h, which proved the feasibility of this sensor in food safety characterization of poly(vinyl alcohol)/sodium alginate hydrogel with high tough-
monitoring. ness and electric conductivity. Carbohydrate Polymers, 186, 377–383.

8
L. Ding, et al. Carbohydrate Polymers 233 (2020) 115859

Kassal, P., Zubak, M., Scheipl, G., Mohr, G. J., Steinberg, M. D., & Murković Steinberg, I. Pacquit, A., Frisby, J., Diamond, D., Lau, K., Farrell, A., Quilty, B., & Diamond, D. (2007).
(2017). Smart bandage with wireless connectivity for optical monitoring of pH. Development of a smart packaging for the monitoring of fish spoilage. Food Chemistry,
Sensors and Actuators B, Chemical, 246, 455–460. 102(2), 466–470.
Kuswandi, B., Jayus Restyana, A., Abdullah, A., Heng, L. Y., & Ahmad, M. (2012). A novel Pereira, V. A., de Arruda, I. N. Q., & Stefani, R. (2015). Active chitosan/PVA films with
colorimetric food package label for fish spoilage based on polyaniline film. Food anthocyanins from brassica oleraceae (red cabbage) as time–temperature indicators
Control, 25(1), 184–189. for application in intelligent food packaging. Food Hydrocolloids, 43, 180–188.
Liang, T., Sun, G., Cao, L., Li, J., & Wang, L. (2018). Rheological behavior of film-forming Rol, F., Belgacem, M. N., Gandini, A., & Bras, J. (2019). Recent advances in surface-
solutions and film properties from Artemisia sphaerocephala Krasch. Gum and purple modified cellulose nanofibrils. Progress in Polymer Science, 88, 241–264.
onion peel extract. Food Hydrocolloids, 82, 124–134. Shi, X., Zhang, L., Cai, J., Cheng, G., Zhang, H., Li, J., & Wang, X. (2011). A facile con-
Liu, P., Guo, X., Nan, F., Duan, Y., & Zhang, J. (2017). Modifying mechanical, optical struction of supramolecular complex from polyaniline and cellulose in aqueous
properties and thermal processability of iridescent cellulose nanocrystal films using system. Macromolecules, 44(12), 4565–4568.
ionic liquid. ACS Applied Materials & Interfaces, 9(3), 3085–3092. Silva-Pereira, M. C., Teixeira, J. A., Pereira-Júnior, V. A., & Stefani, R. (2015). Chitosan/
Liu, S., Zhang, L., Sun, Y., Lin, Y., Zhang, X., & Nishiyama, Y. (2009). Supramolecular corn starch blend films with extract from brassica oleraceae (red cabbage) as a visual
structure and properties of high strength regenerated cellulose films. Macromolecular indicator of fish deterioration. LWT - Food Science and Technology, 61(1), 258–262.
Bioscience, 9(1), 29–35. Wang, S., Lu, A., & Zhang, L. (2016). Recent advances in regenerated cellulose materials.
Ma, Q., Du, L., & Wang, L. (2017). Tara gum/polyvinyl alcohol-based colorimetric NH3 Progress in Polymer Science, 53, 169–206.
indicator films incorporating curcumin for intelligent packaging. Sensors and Wang, W., Fu, S., Leu, S.-Y., & Dong, C. (2018). A nano-ink for gel pens based on scalable
Actuators B, Chemical, 244, 759–766. CNC preparation. Cellulose, 25(11), 6465–6478.
Ma, Q., Du, L., Yang, Y., & Wang, L. (2017). Rheology of film-forming solutions and Wei, Y. C., Cheng, C. H., Ho, Y. C., Tsai, M. L., & Mi, F. L. (2017). Active gellan gum/
physical properties of tara gum film reinforced with polyvinyl alcohol (PVA). Food purple sweet potato composite films capable of monitoring pH variations. Food
Hydrocolloids, 63, 677–684. Hydrocolloids, 69, 491–502.
Mohr, G. J., Werner, T., & Wolfbeis, O. S. (1994). Synthesis of reactive vinylsulphonyl azo Xiong, R., Han, Y., Wang, Y., Zhang, W., Zhang, X., & Lu, C. (2014). Flexible, highly
dyes for application in optical pH sensing. Dyes and Pigments, 24, 223–240. transparent and iridescent all-cellulose hybrid nanopaper with enhanced mechanical
Niu, X., Liu, Y., Fang, G., Huang, C., Rojas, O. J., & Pan, H. (2018). Highly transparent, strength and writable surface. Carbohydrate Polymers, 113, 264–271.
strong, and flexible films with modified cellulose nanofiber bearing UV shielding Youssef, A. M., & El-Sayed, S. M. (2018). Bionanocomposites materials for food packaging
property. Biomacromolecules, 19(12), 4565–4575. applications: Concepts and future outlook. Carbohydrate Polymers, 193, 19–27.
Nopwinyuwong, A., Trevanich, S., & Suppakul, P. (2010). Development of a novel col- Zhang, H., Hou, A., Xie, K., & Gao, A. (2019). Smart color-changing paper packaging
orimetric indicator label for monitoring freshness of intermediate-moisture dessert sensors with pH sensitive chromophores based on azo-anthraquinone reactive dyes.
spoilage. Talanta, 81(3), 1126–1132. Sensors and Actuators B, Chemical, 286, 362–369.
Önal, A., Tekkeli, S. E. K., & Önal, C. (2013). A review of the liquid chromatographic Zhang, X., Lu, S., & Chen, X. (2014). A visual pH sensing film using natural dyes from
methods for the determination of biogenic amines in foods. Food Chemistry, 138(1), bauhinia blakeana Dunn. Sensors and Actuators B, Chemical, 198, 268–273.
509–515.