- Email: [email protected]
Nonenzymatic glucose sensing by CuO nanoparticles decorated nitrogen-doped graphene aerogel
Accepted Manuscript Nonenzymatic glucose sensing by CuO nanoparticles decorated nitrogen-doped graphene aerogel
Jing Yang, Wensheng Tan, Chuanxiang Chen, Yongxin Tao, Yong Qin, Yong Kong PII: DOI: Reference:
S0928-4931(17)31025-1 doi: 10.1016/j.msec.2017.04.097 MSC 7920
To appear in:
Materials Science & Engineering C
Received date: Revised date: Accepted date:
17 March 2017 15 April 2017 16 April 2017
Please cite this article as: Jing Yang, Wensheng Tan, Chuanxiang Chen, Yongxin Tao, Yong Qin, Yong Kong , Nonenzymatic glucose sensing by CuO nanoparticles decorated nitrogen-doped graphene aerogel. The address for the corresponding author was captured as affiliation for all authors. Please check if appropriate. Msc(2017), doi: 10.1016/ j.msec.2017.04.097
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT Nonenzymatic glucose sensing by CuO nanoparticles decorated nitrogen-doped graphene aerogel Jing Yanga, Wensheng Tanb, Chuanxiang Chenc, Yongxin Taoa, Yong Qina, Yong Konga,* Advanced Catalysis and Green Manufacturing Collaborative Innovation Center, School of Petrochemical
T
a
b
CR IP
Engineering, Changzhou University, Changzhou 213164, China
Changzhou Key Laboratory of Large Plastic Parts Intelligence Manufacturing, Changzhou College of Information
c
US
Technology, Changzhou 213164, China
School of Environmental and Chemical Engineering, Jiangsu University of Science and Technology, Zhenjiang
1
AC
CE
PT
ED
M
AN
212003, China
Corresponding author. Tel.: +86 519 86330253, fax: +86 519 86330167. Email address: [email protected] (Y. Kong).
1
ACCEPTED MANUSCRIPT Abstract CuO nanoparticles decorated N-doped graphene aerogel (NGA-CuO) was facilely synthesized via a mild hydrothermal method followed by freeze-drying and calcination, which was characterized by TEM,
T
FT-IR, XPS, XRD and electrochemical impedance spectroscopy (EIS). The obtained NGA-CuO was used
CR IP
for the construction of a nonenzymatic sensing platform for glucose, exhibiting wide linear range, low detection limit, high sensitivity, reproducibility, selectivity and stability. The excellent analytical
US
performances of the NGA-CuO based glucose sensor might be attributed to the synergistic effect of CuO nanoparticles and N-doped graphene aerogel (NGA). The practical applications of the proposed sensor was
AN
verified by determining glucose concentrations in five human serum samples, and the obtained results were
Nonenzymatic; Glucose sensor; CuO nanoparticles; N-doped graphene aerogel; Human
PT
Keywords:
ED
M
quite comparable to those measured on the standard clinical instrument.
AC
CE
serum samples
2
ACCEPTED MANUSCRIPT 1.
Introduction Reliable and fast monitoring of glucose is of great importance for clinical diagnostics of diabetes
mellitus and for food and pharmaceutical quality control [1‒3]. Electrochemical glucose sensors,
T
especially amperometric sensors, have drawn increasing attention owing to exceptional advantages
CR IP
such as fast response, high sensitivity and reliability, good selectivity and low cost [4,5]. In fact, since Clark and Lyons reported the first enzyme glucose sensor in 1962 [6], the glucose oxidase
US
(GOD)-based glucose sensors have gained popularity due to their high sensitivity and specificity to glucose [3,7,8]. However, there are still many obstacles and challenges related to the achievement of
AN
highly reliable and stable glucose monitoring owing to the drawbacks associated with GOD such as
M
high sensitivity to the environmental conditions, cumbersome procedures for enzyme immobilization
ED
and high cost [9,10]. Therefore, the development of enzyme-free (nonenzymatic) glucose sensors as a promising alternative to the traditional enzyme glucose sensors has drawn tremendous academic and
PT
commercial interest.
CE
The past years have witnessed the extensive research of noble metals (Au, Pt, Ag, Pb, and Pd) and
AC
their alloys (Pt–Pd, Au–Ag, Pt–Pb, and so on) in the fabrication of non-enzymatic glucose sensors [11‒15]. However, the high cost of noble metals seriously limits the practical applications of these glucose sensors [16]. Therefore, the development of active and cost-effective catalysts for nonenzymatic glucose sensing still remains an urgent task. Recently, metal oxides, such as CuO, Co3O4, NiO, Fe2O3 and WO3, have been proposed for the construction of enzyme-free glucose sensors [17‒19]. Among these affordable metal oxides, CuO is of particular importance due to its good electrocatalytic activity, high chemical stability and non-toxicity [20]. As a kind of p-type transition metal oxides with 3
ACCEPTED MANUSCRIPT a narrow band gap of 1.2 eV [21], CuO has been widely used as electrode materials in electrochemical sensors [22], supercapacitors [23] and batteries [24,25]. Graphene, a two-dimensional sp2-hybridized carbon, has drawn enormous attention owing to its
T
exceptional mechanical, thermal and electrical properties. Owing to the comparable atomic size to C
CR IP
atom and high electronegativity, N atom has been considered as an ideal dopant for graphene [26]. The incorporation of N into graphene lattice can not only greatly increase the conductivity due to the
US
additional electron density donated by the doped N atom, but also enhance surface chemical activity and electron transfer rate [27‒29]. To the best of our knowledge, little attention has been paid to the
AN
nonenzymatic glucose sensing with the hybrids of CuO and N-doped graphene.
M
In this study, we report on a novel nonenzymatic glucose sensor constructed by decorating N-doped
ED
graphene with CuO nanoparticles. Noted that irreversible aggregation owing to van der Waals force and π−π conjugation between graphene layers usually leads to the overlapping of graphene nanosheets
PT
[30], which severely deteriorates the surface area and analytical performances of graphene [31]. Here,
CE
N-doped graphene aerogel (NGA) was prepared as the support of CuO nanoparticles, since graphene
AC
aerogel synthesized via the feasible self-assembly hydrothermal approach is recognized as a class of three-dimensional (3D) porous graphene architectures [32,33]. The obtained CuO nanoparticles decorated N-doped graphene aerogel (NGA-CuO) was then used for nonenzymatic glucose sensing in human serum samples, exhibiting excellent analytical performances. 2.
Experimental
2.1. Reagents and apparatus
4
ACCEPTED MANUSCRIPT Natural graphite powder (99.95%, 325 mesh), copper sulfate pentahydrate (CuSO4·5H2O), urea and other chemicals not mentioned were purchased from Sinopharm Chemical Reagent Co., Ltd (SCRC, China). D-(+)-Glucose (99.5%) was received from Aladdin Chemistry Co., Ltd (Shanghai, China). All
T
the reagents were of analytical grade and used as received. All aqueous solutions were prepared using
CR IP
ultrapure water with a resistivity of 18.2 MΩ (Milli-Q, Millipore). The human serum samples were obtained from Changzhou No. 2 People’s Hospital, and the use of them was in line with the
US
institutional guidelines.
The morphologies of the as-prepared NGA-CuO were characterized on a JEM-2100 transmission
AN
electron microscope (TEM, JEOL). Fourier transform infrared (FT-IR) spectra of NGA-CuO, CuO and
M
NGA were recorded on a Nicolet iS5 spectrometer (Thermo Fisher Scientific, USA). X-ray
ED
photoelectron spectroscopy (XPS) was measured by a Thermo Fisher ESCALAB 250Xi X-ray photoelectron spectrometer. X-ray powder diffraction (XRD) analysis of graphene oxide (GO), NGA,
PT
CuO and NGA-CuO was carried out on a Rigaku D/max2500PC X-ray powder diffractometer (XRD).
CE
All electrochemical experiments including cyclic voltammetry (CV), electrochemical impedance
AC
spectroscopy (EIS) and chronoamperometry were performed at room temperature on a CHI-660D electrochemical workstation in a conventional three-electrode system consisting of NGA-CuO modified glassy carbon electrode (GCE, φ = 3 mm) as the working electrode, a platinum foil (10 × 5 mm) electrode as the counter electrode, and a saturated calomel electrode (SCE) as the reference electrode, respectively. 2.2. Preparation of NGA-CuO modified GCE First, GO was prepared from natural graphite by the modified Hummers method [34], and 100 mg 5
ACCEPTED MANUSCRIPT of GO was dispersed in 50 mL of ultrapure water, producing a yellow-brown dispersion (2 mg mL−1). CuSO4·5H2O (60 mg), urea (30 mg), NaOH (50 mg) and NaH2PO4·2H2O (78 mg) were added into the GO dispersion, and the mixture was continuously stirred under ultrasonication for 30 min. Next, the
T
mixture was transferred to a Teflon-lined autoclave and then hydrothermally treated at 180 °C for 12 h.
CR IP
After that, the autoclave was naturally cooled to room temperature, and the produced hybrid hydrogel was washed thoroughly with ultrapure water to remove any impurities. Finally, the N-doped hybrid
US
hydrogel was freeze-dried at −52 °C in a vacuum lyophilizer for 12 h and then calcinated in a muffle furnace at 300 °C for 4 h, generating the wanted NGA-CuO. For control experiments, CuO
AN
nanoparticles and un-doped graphene aerogel decorated with CuO nanoparticles (GA-CuO) were also
M
prepared according to the same procedures except for the addition of GO as the support and the
ED
addition of urea as the source of N, respectively. The schematic illustration showing the preparation of
PT
the NGA-CuO is shown in Figure 1.
< Figure 1>
CE
NGA-CuO modified GCE was prepared by a simple casting method. Before casting, the GCE was
AC
polished with 0.05 μm alumina slurries, and then sonicated in a HNO3/water mixture (v/v, 1:1), ethanol and ultrapure water for 4 min each. Next, 10 μL of the NGA-CuO dispersion (2 mg mL-1, dispersed in ultrapure water) was cast onto the surface of the pretreated GCE and then dried in air to obtain the NGA-CuO modified GCE. For a comparison, CuO and GA-CuO modified GCEs were also prepared by casting the same amount (2 mg mL-1) of catalyst (CuO or GA-CuO) onto the surface of GCE. 2.3. Electrochemical measurements EIS measurements were carried out at 0.2 V in 0.1 M KCl containing 10 mM [Fe(CN)6]3-/4-, with a 6
ACCEPTED MANUSCRIPT disturbance potential of 5 mV and a frequency range from 1 MHz to 0.01 Hz. The equivalent circuit of the Nyquist plots was simulated by the ZSimpWin software. CV testing was performed in 0.1 M NaOH with and without 5 mM glucose at a scan rate of 50 mV s-1. Chronoamperometric measurements were
T
implemented at a constant potential of +0.4 V in 20 mL of 0.1 M N2-saturated NaOH by injecting
3.
CR IP
glucose or other interfering species under continuous stirring. Results and discussion
US
3.1. Characterization of NGA-CuO 3.1.1. TEM images of NGA-CuO
AN
Figure 2 shows the typical TEM and high-resolution TEM (HRTEM) images of the NGA-CuO. As
M
can be seen, CuO nanoparticles are well dispersed on the single or few layer structured NGA, and
ED
hardly any free CuO particles are dispersed out of the NGA sheets. The strong anchoring of CuO on the NGA support might be attributed to the 3D structure of NGA, which is beneficial for the fine
PT
accommodation of CuO nanoparticles. In addition, the lattice spacing marked in the HRTEM image is
< Figure 2>
AC
CE
0.24 nm, which could be indexed to the (111) crystallographic plane of CuO [35].
3.1.2. FT-IR spectra of NGA-CuO, CuO and NGA The FT-IR spectra of CuO, NGA and NGA-CuO are shown in Figure 3. For CuO, the peak centered at 511 cm-1 is ascribed to the vibrations of the Cu–O functional group [36]. For NGA, the broad peak around 3433 cm-1 is due to the stretching vibrations of the O–H bonds from the residual oxygen-containing groups, and the peaks at 1574 and 1181 cm-1 are associated with the deformation vibrations of the C=N and C−N bonds, respectively [37], indicating the successful doping of nitrogen 7
ACCEPTED MANUSCRIPT to the graphene lattice. These characteristic peaks of both CuO and NGA can still be observed on the spectrum of NGA-CuO with a slight shift, which implies the integration of CuO nanoparticles to the NGA support.
T
< Figure 3>
CR IP
3.1.3. XPS surveys of NGA-CuO
XPS surveys of NGA-CuO are studied to gain a deep understanding of its chemical composition. As can be seen from Figure 4A, the full XPS spectrum of NGA-CuO shows the presence of C 1s (284.1
US
eV), N 1s (398.5 eV), O 1s (531.1 eV) and Cu 2p (933.8 eV). The existence of N 1s peak implies that
AN
nitrogen could be successfully introduced into the graphene lattice by using urea as the source of N.
M
The de-convoluted C 1s, N 1s and Cu 2p spectra are further investigated. The C 1s spectrum reveals three peaks at 284.1, 285.8, and 288.2 eV (Figure 4B), corresponding to C−C, C−N, and C=O species,
ED
respectively [38,39]. Figure 4C shows the N 1s deconvolution spectrum, in which three peaks
PT
attributed to pyridinic N (398.0 eV), pyrrolic N (399.3 eV), and graphitic N (401.0 eV) are observed.
CE
Compared with graphitic N, the high intensity of pyridinic N and pyrrolic N indicates pyridinic N and pyrrolic N are the dominant configurations of N. It is noteworthy that pyridinic N possesses a pair of
AC
p-electrons, which can effectively change the surface electron donor/acceptor properties and enhance the electrical conductivity [40,41]. The Cu 2p spectrum can be fitted into four peaks (Figure 4D), which correspond to Cu 2p3/2 (933.9 eV) and its satellite peak (942.3 eV), Cu 2p1/2 (954.1 eV) and its satellite peak (962.7 eV), respectively [42], and this result further verifies the integration of CuO particles to the NGA. < Figure 4> 8
ACCEPTED MANUSCRIPT 3.1.4. XRD pattern of NGA-CuO Figure 5 shows the XRD patterns of CuO, NGA, GO and NGA-CuO. For CuO, the peaks at 32.6o, 35.5o, 38.7o, 48.7o, 53.4o, 58.3o, 61.4o, 66.1o, 68.0 o, 72.3o, and 75.2o are attributed to the (110), (-111), (111), (-202), (020), (202), (-113), (-311), (220), (311), and (-222) crystal planes, respectively (JCPDS,
CR IP
T
No. 41-0254) [43]. For NGA, a broad peak centered around 25.2o is observed, which can be assigned to the (002) plane of NGA. In addition, a small peak indexed to the (100) plane appears at around 2θ = 43.4°, suggesting that not all the six carbon atoms of a closed ring in the honeycomb structure are sp2
US
hybridized [44]. Noted that no peak at around 2θ = 9.8o corresponding to the (002) plane of GO is
AN
observed on the pattern of NGA, indicating that GO is completely reduced to NGA. The diffraction peaks of CuO and NGA appear on the XRD pattern of NGA-CuO, demonstrating the successful
M
incorporation of CuO and NGA.
ED
< Figure 5>
PT
3.2. EIS testing of different electrodes
CE
The impedance changes occurring at the electrode-solution interfaces can be known by EIS testing. Figure 6 shows the Nyquist plots of different electrodes. The impedance semicircle in high frequency
AC
corresponds to the interfacial charge transfer resistance (Rct), which is the most directive and sensitive factor showing the changes at the electrode-solution interfaces [45]. As can be seen, the CuO/GCE exhibits the largest Rct (1436 Ω), and it might be attributed to the porous structure of CuO acting as a barrier for the electrochemical process and hindering the access of the redox probe, [Fe(CN)6]3-/4-, to the electrode surface [46]. Compared with bare GCE and CuO/GCE, Rct is greatly decreased at the GA-CuO/GCE (75 Ω) due to the excellent electrical conductivity of GA. Noted that the value of Rct is
9
ACCEPTED MANUSCRIPT further decreased to 64 Ω at the NGA-CuO/GCE, and the smallest Rct at the NGA-CuO/GCE is presumably due to the contributions of the additional electron density donated by the N atoms.
T
3.3. Cyclic voltammograms of different electrodes
CR IP
Figure 7 shows the cyclic voltammograms (CVs) of bare GCE, CuO/GCE, GA-CuO/GCE and NGA-CuO/GCE in 0.1 M NaOH without and with 5 mM glucose at a scan rate of 50 mV s−1. As shown
US
in Figure 7A, the bare GCE does not display any redox peak with and without addition of glucose, indicating that direct oxidation of glucose can not occur for bare GCE. The CuO/GCE shows a small
AN
and broad reduction peak at around 0.6 V (Figure 7B), corresponding to the Cu(II)/Cu(III) redox
M
couple in alkaline media [47], however, the oxidation peak of Cu(II)/Cu(III) is overlapped by the
ED
oxidation process arising from water breakdown. With the addition of glucose, a small oxidation peak appears at around 0.4 V accompanied by the decrease in reduction peak current, which is due to the
PT
irreversible oxidation of glucose. During the oxidation of glucose to gluconolactone, Cu(III) is
CE
consumed as the oxidant and reduced to Cu(II), and therefore decreased reduction peak current at 0.6 V
AC
is observed. For GA-CuO/GCE (Figure 7C), a similar CV behavior as CuO/GCE but a higher oxidation peak current occurs (21 μA versus 15 μA), demonstrating a higher electrocatalytic activity of GA-CuO for the direct oxidation of glucose. Compared with CuO/GCE, the enhanced activity of GA-CuO/GCE might be attributed to the fact that the large surface area of GA provides more active sites for the immobilization of CuO particles, which play a crucial role in the nonenzymatic glucose sensing. The peak current of glucose oxidation is further increased to 32 μA for the NGA-CuO/GCE (Figure 7D), since the doped N atoms can not only increase the conductivity due to the additional electron density 10
ACCEPTED MANUSCRIPT but also enhance surface chemical activity and electron transfer rate [27‒29].
Jing Yang, Wensheng Tan, Chuanxiang Chen, Yongxin Tao, Yong Qin, Yong Kong PII: DOI: Reference:
S0928-4931(17)31025-1 doi: 10.1016/j.msec.2017.04.097 MSC 7920
To appear in:
Materials Science & Engineering C
Received date: Revised date: Accepted date:
17 March 2017 15 April 2017 16 April 2017
Please cite this article as: Jing Yang, Wensheng Tan, Chuanxiang Chen, Yongxin Tao, Yong Qin, Yong Kong , Nonenzymatic glucose sensing by CuO nanoparticles decorated nitrogen-doped graphene aerogel. The address for the corresponding author was captured as affiliation for all authors. Please check if appropriate. Msc(2017), doi: 10.1016/ j.msec.2017.04.097
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT Nonenzymatic glucose sensing by CuO nanoparticles decorated nitrogen-doped graphene aerogel Jing Yanga, Wensheng Tanb, Chuanxiang Chenc, Yongxin Taoa, Yong Qina, Yong Konga,* Advanced Catalysis and Green Manufacturing Collaborative Innovation Center, School of Petrochemical
T
a
b
CR IP
Engineering, Changzhou University, Changzhou 213164, China
Changzhou Key Laboratory of Large Plastic Parts Intelligence Manufacturing, Changzhou College of Information
c
US
Technology, Changzhou 213164, China
School of Environmental and Chemical Engineering, Jiangsu University of Science and Technology, Zhenjiang
1
AC
CE
PT
ED
M
AN
212003, China
Corresponding author. Tel.: +86 519 86330253, fax: +86 519 86330167. Email address: [email protected] (Y. Kong).
1
ACCEPTED MANUSCRIPT Abstract CuO nanoparticles decorated N-doped graphene aerogel (NGA-CuO) was facilely synthesized via a mild hydrothermal method followed by freeze-drying and calcination, which was characterized by TEM,
T
FT-IR, XPS, XRD and electrochemical impedance spectroscopy (EIS). The obtained NGA-CuO was used
CR IP
for the construction of a nonenzymatic sensing platform for glucose, exhibiting wide linear range, low detection limit, high sensitivity, reproducibility, selectivity and stability. The excellent analytical
US
performances of the NGA-CuO based glucose sensor might be attributed to the synergistic effect of CuO nanoparticles and N-doped graphene aerogel (NGA). The practical applications of the proposed sensor was
AN
verified by determining glucose concentrations in five human serum samples, and the obtained results were
Nonenzymatic; Glucose sensor; CuO nanoparticles; N-doped graphene aerogel; Human
PT
Keywords:
ED
M
quite comparable to those measured on the standard clinical instrument.
AC
CE
serum samples
2
ACCEPTED MANUSCRIPT 1.
Introduction Reliable and fast monitoring of glucose is of great importance for clinical diagnostics of diabetes
mellitus and for food and pharmaceutical quality control [1‒3]. Electrochemical glucose sensors,
T
especially amperometric sensors, have drawn increasing attention owing to exceptional advantages
CR IP
such as fast response, high sensitivity and reliability, good selectivity and low cost [4,5]. In fact, since Clark and Lyons reported the first enzyme glucose sensor in 1962 [6], the glucose oxidase
US
(GOD)-based glucose sensors have gained popularity due to their high sensitivity and specificity to glucose [3,7,8]. However, there are still many obstacles and challenges related to the achievement of
AN
highly reliable and stable glucose monitoring owing to the drawbacks associated with GOD such as
M
high sensitivity to the environmental conditions, cumbersome procedures for enzyme immobilization
ED
and high cost [9,10]. Therefore, the development of enzyme-free (nonenzymatic) glucose sensors as a promising alternative to the traditional enzyme glucose sensors has drawn tremendous academic and
PT
commercial interest.
CE
The past years have witnessed the extensive research of noble metals (Au, Pt, Ag, Pb, and Pd) and
AC
their alloys (Pt–Pd, Au–Ag, Pt–Pb, and so on) in the fabrication of non-enzymatic glucose sensors [11‒15]. However, the high cost of noble metals seriously limits the practical applications of these glucose sensors [16]. Therefore, the development of active and cost-effective catalysts for nonenzymatic glucose sensing still remains an urgent task. Recently, metal oxides, such as CuO, Co3O4, NiO, Fe2O3 and WO3, have been proposed for the construction of enzyme-free glucose sensors [17‒19]. Among these affordable metal oxides, CuO is of particular importance due to its good electrocatalytic activity, high chemical stability and non-toxicity [20]. As a kind of p-type transition metal oxides with 3
ACCEPTED MANUSCRIPT a narrow band gap of 1.2 eV [21], CuO has been widely used as electrode materials in electrochemical sensors [22], supercapacitors [23] and batteries [24,25]. Graphene, a two-dimensional sp2-hybridized carbon, has drawn enormous attention owing to its
T
exceptional mechanical, thermal and electrical properties. Owing to the comparable atomic size to C
CR IP
atom and high electronegativity, N atom has been considered as an ideal dopant for graphene [26]. The incorporation of N into graphene lattice can not only greatly increase the conductivity due to the
US
additional electron density donated by the doped N atom, but also enhance surface chemical activity and electron transfer rate [27‒29]. To the best of our knowledge, little attention has been paid to the
AN
nonenzymatic glucose sensing with the hybrids of CuO and N-doped graphene.
M
In this study, we report on a novel nonenzymatic glucose sensor constructed by decorating N-doped
ED
graphene with CuO nanoparticles. Noted that irreversible aggregation owing to van der Waals force and π−π conjugation between graphene layers usually leads to the overlapping of graphene nanosheets
PT
[30], which severely deteriorates the surface area and analytical performances of graphene [31]. Here,
CE
N-doped graphene aerogel (NGA) was prepared as the support of CuO nanoparticles, since graphene
AC
aerogel synthesized via the feasible self-assembly hydrothermal approach is recognized as a class of three-dimensional (3D) porous graphene architectures [32,33]. The obtained CuO nanoparticles decorated N-doped graphene aerogel (NGA-CuO) was then used for nonenzymatic glucose sensing in human serum samples, exhibiting excellent analytical performances. 2.
Experimental
2.1. Reagents and apparatus
4
ACCEPTED MANUSCRIPT Natural graphite powder (99.95%, 325 mesh), copper sulfate pentahydrate (CuSO4·5H2O), urea and other chemicals not mentioned were purchased from Sinopharm Chemical Reagent Co., Ltd (SCRC, China). D-(+)-Glucose (99.5%) was received from Aladdin Chemistry Co., Ltd (Shanghai, China). All
T
the reagents were of analytical grade and used as received. All aqueous solutions were prepared using
CR IP
ultrapure water with a resistivity of 18.2 MΩ (Milli-Q, Millipore). The human serum samples were obtained from Changzhou No. 2 People’s Hospital, and the use of them was in line with the
US
institutional guidelines.
The morphologies of the as-prepared NGA-CuO were characterized on a JEM-2100 transmission
AN
electron microscope (TEM, JEOL). Fourier transform infrared (FT-IR) spectra of NGA-CuO, CuO and
M
NGA were recorded on a Nicolet iS5 spectrometer (Thermo Fisher Scientific, USA). X-ray
ED
photoelectron spectroscopy (XPS) was measured by a Thermo Fisher ESCALAB 250Xi X-ray photoelectron spectrometer. X-ray powder diffraction (XRD) analysis of graphene oxide (GO), NGA,
PT
CuO and NGA-CuO was carried out on a Rigaku D/max2500PC X-ray powder diffractometer (XRD).
CE
All electrochemical experiments including cyclic voltammetry (CV), electrochemical impedance
AC
spectroscopy (EIS) and chronoamperometry were performed at room temperature on a CHI-660D electrochemical workstation in a conventional three-electrode system consisting of NGA-CuO modified glassy carbon electrode (GCE, φ = 3 mm) as the working electrode, a platinum foil (10 × 5 mm) electrode as the counter electrode, and a saturated calomel electrode (SCE) as the reference electrode, respectively. 2.2. Preparation of NGA-CuO modified GCE First, GO was prepared from natural graphite by the modified Hummers method [34], and 100 mg 5
ACCEPTED MANUSCRIPT of GO was dispersed in 50 mL of ultrapure water, producing a yellow-brown dispersion (2 mg mL−1). CuSO4·5H2O (60 mg), urea (30 mg), NaOH (50 mg) and NaH2PO4·2H2O (78 mg) were added into the GO dispersion, and the mixture was continuously stirred under ultrasonication for 30 min. Next, the
T
mixture was transferred to a Teflon-lined autoclave and then hydrothermally treated at 180 °C for 12 h.
CR IP
After that, the autoclave was naturally cooled to room temperature, and the produced hybrid hydrogel was washed thoroughly with ultrapure water to remove any impurities. Finally, the N-doped hybrid
US
hydrogel was freeze-dried at −52 °C in a vacuum lyophilizer for 12 h and then calcinated in a muffle furnace at 300 °C for 4 h, generating the wanted NGA-CuO. For control experiments, CuO
AN
nanoparticles and un-doped graphene aerogel decorated with CuO nanoparticles (GA-CuO) were also
M
prepared according to the same procedures except for the addition of GO as the support and the
ED
addition of urea as the source of N, respectively. The schematic illustration showing the preparation of
PT
the NGA-CuO is shown in Figure 1.
< Figure 1>
CE
NGA-CuO modified GCE was prepared by a simple casting method. Before casting, the GCE was
AC
polished with 0.05 μm alumina slurries, and then sonicated in a HNO3/water mixture (v/v, 1:1), ethanol and ultrapure water for 4 min each. Next, 10 μL of the NGA-CuO dispersion (2 mg mL-1, dispersed in ultrapure water) was cast onto the surface of the pretreated GCE and then dried in air to obtain the NGA-CuO modified GCE. For a comparison, CuO and GA-CuO modified GCEs were also prepared by casting the same amount (2 mg mL-1) of catalyst (CuO or GA-CuO) onto the surface of GCE. 2.3. Electrochemical measurements EIS measurements were carried out at 0.2 V in 0.1 M KCl containing 10 mM [Fe(CN)6]3-/4-, with a 6
ACCEPTED MANUSCRIPT disturbance potential of 5 mV and a frequency range from 1 MHz to 0.01 Hz. The equivalent circuit of the Nyquist plots was simulated by the ZSimpWin software. CV testing was performed in 0.1 M NaOH with and without 5 mM glucose at a scan rate of 50 mV s-1. Chronoamperometric measurements were
T
implemented at a constant potential of +0.4 V in 20 mL of 0.1 M N2-saturated NaOH by injecting
3.
CR IP
glucose or other interfering species under continuous stirring. Results and discussion
US
3.1. Characterization of NGA-CuO 3.1.1. TEM images of NGA-CuO
AN
Figure 2 shows the typical TEM and high-resolution TEM (HRTEM) images of the NGA-CuO. As
M
can be seen, CuO nanoparticles are well dispersed on the single or few layer structured NGA, and
ED
hardly any free CuO particles are dispersed out of the NGA sheets. The strong anchoring of CuO on the NGA support might be attributed to the 3D structure of NGA, which is beneficial for the fine
PT
accommodation of CuO nanoparticles. In addition, the lattice spacing marked in the HRTEM image is
< Figure 2>
AC
CE
0.24 nm, which could be indexed to the (111) crystallographic plane of CuO [35].
3.1.2. FT-IR spectra of NGA-CuO, CuO and NGA The FT-IR spectra of CuO, NGA and NGA-CuO are shown in Figure 3. For CuO, the peak centered at 511 cm-1 is ascribed to the vibrations of the Cu–O functional group [36]. For NGA, the broad peak around 3433 cm-1 is due to the stretching vibrations of the O–H bonds from the residual oxygen-containing groups, and the peaks at 1574 and 1181 cm-1 are associated with the deformation vibrations of the C=N and C−N bonds, respectively [37], indicating the successful doping of nitrogen 7
ACCEPTED MANUSCRIPT to the graphene lattice. These characteristic peaks of both CuO and NGA can still be observed on the spectrum of NGA-CuO with a slight shift, which implies the integration of CuO nanoparticles to the NGA support.
T
< Figure 3>
CR IP
3.1.3. XPS surveys of NGA-CuO
XPS surveys of NGA-CuO are studied to gain a deep understanding of its chemical composition. As can be seen from Figure 4A, the full XPS spectrum of NGA-CuO shows the presence of C 1s (284.1
US
eV), N 1s (398.5 eV), O 1s (531.1 eV) and Cu 2p (933.8 eV). The existence of N 1s peak implies that
AN
nitrogen could be successfully introduced into the graphene lattice by using urea as the source of N.
M
The de-convoluted C 1s, N 1s and Cu 2p spectra are further investigated. The C 1s spectrum reveals three peaks at 284.1, 285.8, and 288.2 eV (Figure 4B), corresponding to C−C, C−N, and C=O species,
ED
respectively [38,39]. Figure 4C shows the N 1s deconvolution spectrum, in which three peaks
PT
attributed to pyridinic N (398.0 eV), pyrrolic N (399.3 eV), and graphitic N (401.0 eV) are observed.
CE
Compared with graphitic N, the high intensity of pyridinic N and pyrrolic N indicates pyridinic N and pyrrolic N are the dominant configurations of N. It is noteworthy that pyridinic N possesses a pair of
AC
p-electrons, which can effectively change the surface electron donor/acceptor properties and enhance the electrical conductivity [40,41]. The Cu 2p spectrum can be fitted into four peaks (Figure 4D), which correspond to Cu 2p3/2 (933.9 eV) and its satellite peak (942.3 eV), Cu 2p1/2 (954.1 eV) and its satellite peak (962.7 eV), respectively [42], and this result further verifies the integration of CuO particles to the NGA. < Figure 4> 8
ACCEPTED MANUSCRIPT 3.1.4. XRD pattern of NGA-CuO Figure 5 shows the XRD patterns of CuO, NGA, GO and NGA-CuO. For CuO, the peaks at 32.6o, 35.5o, 38.7o, 48.7o, 53.4o, 58.3o, 61.4o, 66.1o, 68.0 o, 72.3o, and 75.2o are attributed to the (110), (-111), (111), (-202), (020), (202), (-113), (-311), (220), (311), and (-222) crystal planes, respectively (JCPDS,
CR IP
T
No. 41-0254) [43]. For NGA, a broad peak centered around 25.2o is observed, which can be assigned to the (002) plane of NGA. In addition, a small peak indexed to the (100) plane appears at around 2θ = 43.4°, suggesting that not all the six carbon atoms of a closed ring in the honeycomb structure are sp2
US
hybridized [44]. Noted that no peak at around 2θ = 9.8o corresponding to the (002) plane of GO is
AN
observed on the pattern of NGA, indicating that GO is completely reduced to NGA. The diffraction peaks of CuO and NGA appear on the XRD pattern of NGA-CuO, demonstrating the successful
M
incorporation of CuO and NGA.
ED
< Figure 5>
PT
3.2. EIS testing of different electrodes
CE
The impedance changes occurring at the electrode-solution interfaces can be known by EIS testing. Figure 6 shows the Nyquist plots of different electrodes. The impedance semicircle in high frequency
AC
corresponds to the interfacial charge transfer resistance (Rct), which is the most directive and sensitive factor showing the changes at the electrode-solution interfaces [45]. As can be seen, the CuO/GCE exhibits the largest Rct (1436 Ω), and it might be attributed to the porous structure of CuO acting as a barrier for the electrochemical process and hindering the access of the redox probe, [Fe(CN)6]3-/4-, to the electrode surface [46]. Compared with bare GCE and CuO/GCE, Rct is greatly decreased at the GA-CuO/GCE (75 Ω) due to the excellent electrical conductivity of GA. Noted that the value of Rct is
9
ACCEPTED MANUSCRIPT further decreased to 64 Ω at the NGA-CuO/GCE, and the smallest Rct at the NGA-CuO/GCE is presumably due to the contributions of the additional electron density donated by the N atoms.
T
3.3. Cyclic voltammograms of different electrodes
CR IP
Figure 7 shows the cyclic voltammograms (CVs) of bare GCE, CuO/GCE, GA-CuO/GCE and NGA-CuO/GCE in 0.1 M NaOH without and with 5 mM glucose at a scan rate of 50 mV s−1. As shown
US
in Figure 7A, the bare GCE does not display any redox peak with and without addition of glucose, indicating that direct oxidation of glucose can not occur for bare GCE. The CuO/GCE shows a small
AN
and broad reduction peak at around 0.6 V (Figure 7B), corresponding to the Cu(II)/Cu(III) redox
M
couple in alkaline media [47], however, the oxidation peak of Cu(II)/Cu(III) is overlapped by the
ED
oxidation process arising from water breakdown. With the addition of glucose, a small oxidation peak appears at around 0.4 V accompanied by the decrease in reduction peak current, which is due to the
PT
irreversible oxidation of glucose. During the oxidation of glucose to gluconolactone, Cu(III) is
CE
consumed as the oxidant and reduced to Cu(II), and therefore decreased reduction peak current at 0.6 V
AC
is observed. For GA-CuO/GCE (Figure 7C), a similar CV behavior as CuO/GCE but a higher oxidation peak current occurs (21 μA versus 15 μA), demonstrating a higher electrocatalytic activity of GA-CuO for the direct oxidation of glucose. Compared with CuO/GCE, the enhanced activity of GA-CuO/GCE might be attributed to the fact that the large surface area of GA provides more active sites for the immobilization of CuO particles, which play a crucial role in the nonenzymatic glucose sensing. The peak current of glucose oxidation is further increased to 32 μA for the NGA-CuO/GCE (Figure 7D), since the doped N atoms can not only increase the conductivity due to the additional electron density 10
ACCEPTED MANUSCRIPT but also enhance surface chemical activity and electron transfer rate [27‒29].
3.4. Amperometric response of the NGA-CuO nonenzymatic sensor
T
Figure 8A shows the typical chronoamperometric responses of NGA-CuO, GA-CuO and CuO
CR IP
based sensors to successive additions of glucose into stirred and N2-saturated 0.1 M NaOH at a potential of +0.4 V. All the three sensors respond quickly to the continuous addition of glucose and
US
reach to the steady-state current within a short time. As expected, the NGA-CuO/GCE shows much better sensing performance for glucose than GA-CuO/GCE and CuO/GCE, agreeing well with the
AN
results of CV, and the detection limit of the NGA-CuO/GCE is calculated to be 2.7 μM (signal/noise =
M
3) according to the method proposed by Li et al [48]. The calibration curves of the three sensors are
ED
shown in Figure 8B, and it is shown that the current responses of these sensors exhibit a linear dependence on glucose concentration from 0.01 to 6.75 mM. According to the known electrode surface
PT
area (0.07 cm2), the sensitivity of the NGA-CuO/GCE is estimated to be 223.1 μA mM−1 cm−2, which
CE
is much higher than those of CuO/GCE (26.9 μA mM−1 cm−2) and GA-CuO/GCE (69.7 μA mM−1
AC
cm−2). It is noteworthy that the sensitivity of the proposed NGA-CuO is even superior to those of other reported nonenzymatic glucose sensors listed in Table 1. Obviously, the excellent analytical performances of the NGA-CuO are attributed to the synergistic effect of NGA and CuO particles. 3.5. Anti-interference, reproducibility and stability study of the NGA-CuO nonenzymatic sensor Anti-interference ability, reproducibility and stability are of great importance for practical 11
T
Figure 8A shows the typical chronoamperometric responses of NGA-CuO, GA-CuO and CuO
CR IP
based sensors to successive additions of glucose into stirred and N2-saturated 0.1 M NaOH at a potential of +0.4 V. All the three sensors respond quickly to the continuous addition of glucose and
US
reach to the steady-state current within a short time. As expected, the NGA-CuO/GCE shows much better sensing performance for glucose than GA-CuO/GCE and CuO/GCE, agreeing well with the
AN
results of CV, and the detection limit of the NGA-CuO/GCE is calculated to be 2.7 μM (signal/noise =
M
3) according to the method proposed by Li et al [48]. The calibration curves of the three sensors are
ED
shown in Figure 8B, and it is shown that the current responses of these sensors exhibit a linear dependence on glucose concentration from 0.01 to 6.75 mM. According to the known electrode surface
PT
area (0.07 cm2), the sensitivity of the NGA-CuO/GCE is estimated to be 223.1 μA mM−1 cm−2, which
CE
is much higher than those of CuO/GCE (26.9 μA mM−1 cm−2) and GA-CuO/GCE (69.7 μA mM−1
AC
cm−2). It is noteworthy that the sensitivity of the proposed NGA-CuO is even superior to those of other reported nonenzymatic glucose sensors listed in Table 1. Obviously, the excellent analytical performances of the NGA-CuO are attributed to the synergistic effect of NGA and CuO particles.