Shell-isolated graphene@Cu nanoparticles on graphene@Cu substrates for the application in SERS

Accepted Manuscript Shell-isolated Graphene@Cu Nanoparticles on Graphene@Cu Substrates for the Application in SERS Cheng Yang, Chao Zhang, Yanyan Huo, Shouzhen Jiang, Hengwei Qiu, Yuanyuan Xu, Xiuhua Li, Baoyuan Man PII:

S0008-6223(15)30451-6

DOI:

10.1016/j.carbon.2015.11.042

Reference:

CARBON 10508

To appear in:

Carbon

Received Date: 13 August 2015 Revised Date:

13 November 2015

Accepted Date: 14 November 2015

Please cite this article as: C. Yang, C. Zhang, Y. Huo, S. Jiang, H. Qiu, Y. Xu, X. Li, B. Man, Shellisolated Graphene@Cu Nanoparticles on Graphene@Cu Substrates for the Application in SERS, Carbon (2015), doi: 10.1016/j.carbon.2015.11.042. 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 1

Shell-isolated

Graphene@Cu

Nanoparticles

2

Substrates for the Application in SERS

on

Graphene@Cu

3

Cheng Yanga, 1, ∗, Chao Zhanga, 1, Yanyan Huoa, Shouzhen Jianga, Hengwei Qiua,

5

Yuanyuan Xua, Xiuhua Lib, Baoyuan Mana, ∗

RI PT

4

6 7

a

8

People’s Republic of China

9

b

SC

M AN U

10

school of Physics and Electronics, Shandong Normal University, Jinan 250014,

Lishan College, Shandong Normal University, Jinan, 250014, People’s Republic of

China

11

TE D

12

∗

AC C

EP

13

Corresponding authors at: school of Physics and Electronics, Shandong Normal

University, Jinan 250014, People’s Republic of China. E-mail addresses: [email protected] (Cheng Yang), [email protected] (Baoyuan Man). 1

These authors contributed equally. 1

ACCEPTED MANUSCRIPT Abstract

2

Graphene layers, which were used to prevent metal-molecule interactions and

3

improve the stability of the metal nanoplasmonics, have drawn a tremendous amount

4

of attention in SERS application. At the same time, Cu substrate is usually etched to

5

transfer the graphene onto the target substrates. The transfer process would always be

6

time-consuming, which severely hinders its future application. Facial fabrication of

7

graphene-SERS (G-SERS) structure with a simple, chemically stable isolating shell

8

on a transfer-free graphene substrate is important. Here, core-shell graphene@Cu

9

nanoparicles (G@CuNPs) have been directly deposited onto the G@Cu substrate to

10

form the all-sided-isolated G@CuNP/G@Cu SERS substrate using the CVD system

11

with a designed process. Raman spectra show that G@CuNPs can dramatically

12

suppress photobleaching and fluorescence of the probe molecules (R6G), resulting in

13

a stronger Raman signal. Electromagnetic mechanism (EM) is the main mechanism of

14

SERS enhanced by its localized surface plasmon resonance (LSPR). Graphene layers

15

will not alter the EM enhancement significantly, but could offer additional chemical

16

mechanism (CM) enhancement to improve the total SERS properties. More

17

importantly, the stability of the G@CuNP/G@Cu substrate is improved, benefiting

18

from that graphene layer can act as the passivation layer to inhibit the surface

19

oxidation of the Cu nanoparticles and Cu substrate.

AC C

EP

TE D

M AN U

SC

RI PT

1

20

2

ACCEPTED MANUSCRIPT 1

1. Introduction Surface-enhanced Raman scattering (SERS) as one of the most important and

3

powerful analytic techniques to probe the chemical interaction between the adsorbing

4

molecules and the surface of some metals has received increasing attention, which

5

could provide signal intensity of the molecules enhanced by orders of magnitude

6

[1-8]. Many different types SERS substrates have been reported in the past few years

7

[9-13]. Among them, the metal nanoplasmonics, such as Ag [3, 6, 9, 12, 13], Au [7, 9,

8

10] and Cu [1, 3] nanoparticles, as SERS-active substrates are of particular interest.

9

Despite considerable efforts, it is still a challenge to achieve ideal SERS substrates

10

with good stability and reproducibility, caused by the signal variations of

11

metal-molecule contact [14, 15]. Furthermore, the lower adsorption capacity of metal

12

nanostructures for some molecules often limits their applications [1, 16]. An ultrathin

13

inert shell, such as SiO2 [17, 18], TiO2 [19], or Al2O3 [17] layers, to isolate metal

14

nanostructures from their surroundings was demonstrated, which must be thin enough

15

to lower the loss of electromagnetic enhancement activity [1, 17].

EP

TE D

M AN U

SC

RI PT

2

Graphene is well-known for its unique electrical performance and the amazing

17

applications in nanoscale electronics [20-25]. There are two key advantages by using

18

the graphene on SERS-active substrate: (1) the 2-dimensional nature of graphene, a

19

favorable atomic surface test bed with the small-distance charge transfer between

20

graphene surface and the adsorbed molecules, makes the Raman signal more reliable

21

and efficient [1, 24-29]. (2) graphene with the large specific surface area of 2630 m2/g

22

can also work as a molecule enricher in SERS-active substrate, which could act as an

AC C

16

3

ACCEPTED MANUSCRIPT 1

excellent adsorbent towards organic molecules, especially the aromatic molecules [1,

2

24-29]. So that, graphene can effectively enhance the Raman signal and reduce the

3

back action noise [26, 27]. Recently, Graphene is used as the ultrathin inert shell to obtain the

5

graphene-noble metallic nanomaterials with a great spectral tuning capability [28, 29].

6

Zhang’s group has shown that few-layer graphene-encapsulated metal nanoparticles

7

hybrid is a promising material for shell isolated SERS [28]. He et al. demonstrated

8

that the gold decorated graphene can serve as a SERS-active substrate for

9

multiplexing detection of DNA [29]. However, the used graphene film was grown

10

using the chemical vapor deposition (CVD) method and should be transferred away

11

from the grown substrate, i.e. Cu foils [3, 28, 29]. The damage or impurities could

12

easily be introduced into graphene during such a transfer process, which may make it

13

difficult to investigate the optical properties of devices and explore the potential

14

applications. In addition, the transfer process would always be time consuming, which

15

severely hinders its future application. Therefore, constructing and studying the

16

optical properties of a hybrid graphene-metal nanostructure without transferring the

17

CVD grown graphene would be greatly desirable.

AC C

EP

TE D

M AN U

SC

RI PT

4

18

4

ACCEPTED MANUSCRIPT Fig. 1-Schematic illustration of Raman experiments on the G@CuNP/G@Cu

2

hybrid structure (a). AFM image of the quasi-periodic array of G@CuNPs on the

3

G@Cu substrate (b). TEM image of the few-layer graphene on the surface of the

4

Cu nanoparticles (c).

RI PT

1

In this work, we provide a direct growth approach to prepare a high

6

performance G@CuNP/G@Cu SERS substrate (shown in Fig. 1a) with G@CuNPs on

7

G@Cu substrate by a CVD method. Relatively uniform G@CuNPs (shown in Fig. 1b)

8

are formed on the surface of the G@Cu foils. Few-layer graphene is easily formed on

9

the surface of the Cu nanoparticles, which can be proved by TEM (JEOL, JEM-2100)

10

results as shown in Fig. 1c. In our proposed growth process, mono-layer graphene on

11

Cu substrates and core-shell few-layer graphene covered Cu nanoparticles are formed

12

sequentially in the same tube to give a all-sided prevention of the metal-molecule

13

interactions. As a chemistry composite mode, this method provides an atomically thin,

14

seamless, and chemically inert net to tightly wrap the Cu nanoparticles and Cu

15

substrate. The application of such new substrates in SERS inherits four advantages: (1)

16

the direct growth mode can make graphene layer tightly wrap the metal nanoparticles,

17

minimize the loss of electromagnetic enhancement activity, and make the graphene

18

cover every place of the CuNPs and Cu substrates, even the narrow gaps of particles

19

(hot spots); (2) formation of the cheaper Cu nanoparticles (instead of Au and Ag) and

20

direct fabrication of the structures (without transferring graphene) will greatly lower

21

the fabrication cost; (3) two enhancement mechanisms of the electromagnetic

22

mechanism (Cu nanoparticles and Cu foil) and the chemical mechanism (graphene)

AC C

EP

TE D

M AN U

SC

5

5

ACCEPTED MANUSCRIPT are combined to give the overall SERS enhancement of the fabricated

2

G@CuNP/G@Cu substrate; (4) the stability of the G@CuNP/G@Cu substrate is

3

improved, benefiting from that graphene layers can act as the passivation layer to

4

inhibit the surface oxidation of the Cu nanoparticles and Cu substrate.

5

2. Experimental

6

2.1 Fabrication of the G@CuNP/G@Cu substrate

EP

8

TE D

M AN U

SC

7

RI PT

1

Fig. 2-Schematic illustration of the growth setup. One side of the Cu foil is

10

attached on the flat side of the quartz substrate (attached zone), another side is

11

suspended in midair with the six quartz rod (suspended zone) (a). Schematic

12

illustration of graphene growth mechanism involving synthesis of graphene on

13

Cu foil and the fabrication of G@CuNPs on G@Cu substrate in the mixture of

14

H2, decomposed CH4 and thermal evaporated Cu vapor (b).

AC C

9

15

The prime feature of the proposed all-sided-isolated substrate fabrication

16

approach includes (1) synthesis of graphene on Cu foil in the mixture H2 and 6

ACCEPTED MANUSCRIPT decomposed CH4 in the beginning, and (2) fabrication of G@CuNPs in the mixture of

2

H2, decomposed CH4 and thermal evaporated Cu vapor. Fig. 2a shows the growth

3

setup of G@CuNPs onto the G@Cu substrate. A strip of Cu foil cleaned by acetic

4

acid surrounding along the tube wall was placed on the designed quartz substrate. Six

5

quartz rods with the diameter of 1 mm and the height of 3 mm are fixed on one side of

6

the quartz substrate, while another side is flat, as shown in Fig. 2a. The size of the Cu

7

foil is same with the quartz substrate. One side of the Cu foil is attached on the flat

8

side of the quartz substrate (attached zone), another side is suspended in midair with

9

the six quartz rods (suspended zone). The quartz substrate, with the attached zone as

SC

M AN U

10

RI PT

1

the upstream zone, was placed in the quartz tube, which was heated up to 1050℃. For the same annealing temperature of 1050℃, the temperature of the solid

12

quartz substrate is higher than that of the surrounding atmosphere of H2 and CH4. So,

13

the Cu foil on the attached zone is easier to sublimate and produce a large number of

14

Cu atoms than that on the suspended zone. The Cu atoms are then transported

15

downstream with gas flow into the suspended zone. Fig. 2b schematically illustrates

16

the working mechanism formation of (1) the graphene on the whole Cu foil and (2)

17

the G@CuNPs on the suspended zone. In the beginning, monolayer graphene is

18

synthesized on the whole surface of the Cu foil. As the growth time is increasing, the

19

G@Cu foil on the attached zone with a higher temperature sublimates and produces

20

more and more Cu atoms. When the Cu atoms attain a certain concentration, these Cu

21

atoms begin to merge with each other, evolving into Cu nanoparticles with a certain

22

size. On the other hand, the Cu atoms are also used as catalysts to decomposes CH4,

AC C

EP

TE D

11

7

ACCEPTED MANUSCRIPT enabling a typically CVD reaction to grow a graphene layer on floating CuNPs. The

2

G@CuNP are then transported downstream with gas flow into the suspended zone.

3

For the G@Cu foil on the suspended zone with a lower temperature, only little Cu

4

atoms are sublimated from the Cu foil. So, the G@CuNPs are deposited onto the

5

G@Cu surface to form the G@CuNP/G@Cu substrates. As a comparison, we also

6

prepared single CuNPs in the mixture of H2 and Ar on the Cu foil.

7

2.2 Theoretical modeling

SC

RI PT

1

Local electric field properties of proposed G@CuNP/G@Cu substrate are

9

obtained by using the finite element method (commercial COMSOL software) to

10

analyze and clarify the enhancement mechanisms of SERS. A plane wave (532nm)

11

irradiates down to G@CuNPs deposited on the G@Cu substrate from the z direction.

12

The electric field is parallel to the x direction. For simplicity, a CuNP sphere with the

13

diameter of 50 nm covered by graphene layer (monolayer ~0.34nm, trilayer ~1nm) is

14

designed on a 100 nm thick Cu film to study the underlying physics of the proposed

15

structure. The refractive index εCu =1.16+i2.6 of the Cu comes from Ref [30]. For the

16

graphene layers, the effective dielectric constant is εG ≈ 5.5585 + i7.4064 at 532 nm,

17

calculated from a Lorentz−Drude model [31, 32].

18

3. Results and discussion

AC C

EP

TE D

M AN U

8

8

SC

RI PT

ACCEPTED MANUSCRIPT

1

Fig.3-Optical microscope images of the G@Cu substrate, G@CuNP/G@Cu and

3

G@CuNP/SiO2 (a)-(c). SEM image of the G@CuNP/G@Cu substrate (d). AFM

4

image of the G@CuNP/G@Cu substrate with a scale of 1×1 um2 (e). AFM image

5

of the flexible G@CuNP/G@Cu substrate with a scale of 5×5 um2 (f).

M AN U

2

Optical microscope, SEM (Zeiss Gemini Ultra-55) and AFM (Bruker

7

Mutimode 8) were used to investigate the surface morphology of G@CuNPs formed

8

on the G@Cu substrate. Fig. 3a and 3b show the optical images of G@Cu substrate

9

without (Fig.3a) and with (Fig. 3b) the formation of the uniform nanoparticles on the

10

surface. It can be seen from Fig. 3b, the G@Cu substrate is covered by a large number

11

of nanoparticles with relatively uniform size. The similar nanoparticles can also be

12

observed on the flat SiO2 substrate (Fig. 3c), which indicates that our proposed

13

method is much efficient for the preparation of relatively uniform CuNPs. To identify

14

the size of nanoparticles, a higher magnification for SEM image was adopted. As

15

shown in Fig. 3d, the average size of these nanoparticles is ~50 nm and the gaps

16

among nanoparticles are very narrow. The relatively uniform G@CuNPs can also be

AC C

EP

TE D

6

9

ACCEPTED MANUSCRIPT identified by the AFM image, shown in the Fig. 3e and 3f. The obtained

2

G@CuNP/G@Cu substrate after the high temperature treatment is flexible and stable.

3

In order to demonstrate the stability of the G@CuNPs, we fold and stretch the

4

G@CuNP/G@Cu substrate several times. It is interesting and important that the

5

relatively uniform G@CuNPs can not be easily damaged by the deformation of the Cu

6

substrate, which can be easily observed by the Fig. 3f and be attributed to the high

7

interaction between G@CuNPs and Cu substrate.

M AN U

SC

RI PT

1

TE D

8

Fig. 4.-The Raman spectra of the G@Cu (upper) and G@CuNP/G@Cu (down)

10

substrates with different laser power (a). G (1580 cm–1) and 2D (2698 cm–1)

11

intensity of the G@CuNP/G@Cu substrates with the different incident laser

12

power (b).

AC C

13

EP

9

As the direct grown graphene layers can not be directly recognized by SEM

14

observations, a conventional Raman measurement was performed on these

15

nanoparticles to demonstrate the existence of graphene with a Horiba HR Evolution

16

800 Raman system (532nm). As shown in Fig. 4a, the D, G and 2D peaks of graphene

17

are clearly observed at ~1360, ~1580 and ~2698 cm–1, respectively, which can be

18

regarded as the fingerprint of graphene [33]. The defect-related D peak is also

10

ACCEPTED MANUSCRIPT detected, which can be due to the existence of the Cu nanoparticles. The Raman

2

spectra of the G@CuNP/G@Cu substrates are distinct from that of G@Cu substrate.

3

Typical Raman signals of few-layer graphene are easily obtained on the

4

G@CuNP/G@Cu substrates, while it is hard to observe the characteristic peaks of the

5

mono-layer graphene on the Cu substrate. As the flat surface of the Cu substrate, the

6

surface plasmons can not be excited by the incident laser. Consequently, in this case,

7

only fluorescence background is observed. Although graphene can effectively

8

suppress the fluorescence background and enhance the Raman signal by the CM, only

9

1-10 enhancement is achieved in this case [32, 34, 35]. Therefore, it is reasonable for

10

the poor enhancement of the characteristic peaks of the G@Cu substrate. Compared to

11

the G@Cu substrate, the G@CuNP/G@Cu has better enhancement, which can be

12

attributed to the EM enhancement introduced by the CuNP [3, 34, 35]. The better

13

enhancement for G@CuNP/G@Cu can be ascribed to the close contact between

14

graphene and CuNP benefitted from the direct growth mode. Fig. 4b shows the G

15

(1580 cm–1) and 2D (2698 cm–1) peaks intensity of the G@CuNP/G@Cu substrates

16

with the different incident laser power. The Raman spectra of the G@CuNP/G@Cu

17

substrates show typical features of few-layer graphene: the intensity ratio of I(G)/I(2D)

18

≥ 1. The observed graphene layers come from the surface of the Cu foil and Cu

19

nanoparticles. To give a clear visualization for the number of the graphene layers on

20

the surface of the Cu nanoparticles, HRTEM analysis is carried out at the edges of

21

G@CuNP. Just as exhibited in Fig. 1c, three dark lines (interlayer spacing ~0.34 nm)

22

are detected, indicative of trilayer graphene [36], which is well consistent with the

AC C

EP

TE D

M AN U

SC

RI PT

1

11

ACCEPTED MANUSCRIPT Raman results. These results prove that the graphene film with a trilayer structure is

2

actually deposited on the surface of the CuNPs.

M AN U

SC

RI PT

1

3

Fig. 5-The Raman spectra and their linear fit calibration curve of the R6G

5

molecules with concentrations from 10−9 to 10−5 M on the G@CuNP/G@Cu (a),

6

(b) and CuNP/Cu (c), (d).

TE D

4

To assess the SERS behavior of the G@CuNP/G@Cu on the same conditions

8

with the conventional Raman measurement, R6G molecules (Little-PA Sciences Co.,

9

Ltd) with concentrations from 10−9 to 10−5 M were used as the probe molecule. As a

10

contrast, the SERS spectra of R6G on the CuNP/Cu were also collected. The primary

11

Raman peaks of R6G are confirmed according to the reported work [37-39]. The

12

peaks at 613 cm-1 is assigned to the in-plane vibration of C-C-C. The peaks at 774 and

13

1185 cm−1 are respectively assigned to the C-H out-of-plane vibration and in-plane

14

vibration. The peaks at 1311, 1360, 1507 and 1645 cm−1 are assigned to the stretching

15

vibration mode of aromatic C-C. As shown in Fig. 5, the SERS spectra can be easily

AC C

EP

7

12

ACCEPTED MANUSCRIPT 1

observed on both of the G@CuNP/G@Cu and CuNP/Cu substrates with a little

2

different of the intensities. Concentration

Concentration

Concentration

Concentration

Substrate

(10-5M)

(10-6 M)

(10-7 M)

(10-8 M)

(10-9 M)

G@CuNP/G@Cu

17850

7911

2869

CuNP/Cu

15583

5824

1566

G@CuNP/G@Cu

7367

3713

1217

CuNP/Cu

6305

G@CuNP/G@Cu

2665

CuNP/Cu

2289

shift

774

1185

M AN U

613

1044

369

802

122

449

167

SC

(cm-1)

RI PT

Concentration

Raman

2752

629

276

83

1741

861

370

101

1023

475

185

69

Table 1. Intensity of the Raman peaks from the G@CuNP/G@Cu and CuNP/Cu

4

substrates

TE D

3

The intensities of SERS spectra on G@CuNP/G@Cu are 1.2-3 times stronger than

6

that of CuNP/Cu (shown in Table 1), caused maybe by the molecule enrichment from

7

graphene and the close contact between graphene and CuNP benefitted from the direct

8

growth mode. These stronger intensities indicate that the G@CuNP/G@Cu is superior

9

to the CuNP/Cu for the enhancement effect. The graphene layers directly deposit on

10

the surface of CuNP and form a tight combination, which can avoid the

11

electromagnetic enhancement decay between metal and analyte and thus lead to large

12

SERS activity. Obviously, the sharp characteristic peaks in SERS spectrum of R6G

13

from G@CuNP/G@Cu exhibit the better signal-to-noise ratio. To represent the

AC C

EP

5

13

ACCEPTED MANUSCRIPT capability of the quantitative detection of R6G, the linear fit calibration curves are

2

illustrated in Fig. 5b and 5d. The reasonable linear response of SERS is observed from

3

5 to 500 nM. Intensity of each peak has a good linear fit calibration curve. However,

4

the linear correlation between SERS spectra intensity and R6G concentration

5

collected on CuNP/Cu is not as good as that collected on G@CuNP/G@Cu substrates.

6

Another crucial parameter for the SERS behavior is the homogeneity for the SERS

7

signals, which can be characterized by the value of standard deviations and the R2.

8

Just as exhibited in Fig. 5b and 5d, the G@CuNP/G@Cu substrate possesses smaller

9

standard deviations and larger R2 for all the peaks at 613，774 and 1185 cm−1

10

compared with that of CuNP/Cu, which can demonstrate the fact that the homogeneity

11

of G@CuNP/G@Cu substrate is perfect.

M AN U

SC

RI PT

1

As we all known, there are two widely accepted mechanism of the

13

electromagnetic mechanism (EM) and the chemical mechanism (CM)) for SERS [37,

14

40]. The enhancement of the EM is roughly proportional to |E|4 [37, 40-42], which is

15

based on the enhancement of the local electromagnetic field upon resonance

16

excitation of localized surface plasmon resonance (LSPR). The CM effect originates

17

from the interaction between molecules and substrates, which is based mainly upon a

18

partially resonant charge transfer between the molecules and the substrate (usually

19

metal) as well as a nonresonant chemical interaction between the ground state of the

20

molecule and metal.

AC C

EP

TE D

12

14

SC

RI PT

ACCEPTED MANUSCRIPT

1

Fig. 6-Electric field distributions of a single Cu nanoparticle (a), Cu nanoparticle

3

covered with monolayer graphene (b), Cu nanoparticle covered with trilayer

4

graphene (d) and 3×3 distributed array particles on Cu substrate with diameter

5

of 50 nm, nanogap of 15 nm (c). Cu nanoparticle covered with three-layer

6

graphene on SiO2 substrate (e).

TE D

M AN U

2

EM is well-known as the main mechanism of SERS for metallic nanostructures

8

(such as Cu nanoparticles) [37]. The local electric field near the metallic nanostructure

9

is obviously enhanced by its LSPRs, which produce an enhancement highly localized

10

to the metal surface. We used the COMSOL Multiphysics software to analyze the

11

local electric field properties of Cu nanoparticles on Cu foil substrate and clarify the

12

enhancement mechanisms of SERS. Fig. 6a, 6c and 6d show the electric field

13

distributions of CuNP/Cu structure. The strong field intensity appears around the

14

nanoparticles because of the local surface plasmon supported by Cu nanoparticles.

15

The electric field intensity of a single particle can enhanced 4 times. The electric field

16

intensity of 3×3 distributed array particles is higher than that of single particle and can

AC C

EP

7

15

ACCEPTED MANUSCRIPT 1

reach to 6.9 times enhancement, which is caused by the coupling effect between Cu

2

nanoparticles. On the other hand, the EM enhancement of the monolayer G@CuNP/G@Cu

4

substrate is a little lower than that of the samples without graphene and higher than

5

that of the samples with trilayer G@CuNPs on G@Cu and SiO2 substrates (shown in

6

Fig. 6c, 6d and 6e). The slightly reduced electric field enhancement is caused by the

7

additional effective dielectric loss of the graphene layer (εG ≈5.5585 + i7.4064 at 532

8

nm), calculated from a Lorentz-Drude model [31, 32]. However, only a small

9

influence on the simulation is to be expected because of the much smaller graphene

10

thickness than its skin depth at 532 nm. So, the SERS enhancement resulting from

11

graphene is believed to be a chemical effect. The similar atomic structure of R6G

12

molecules with that of graphene makes R6G easily oriented themselves parallel to the

13

surface of the graphene due to the π−π stacking [43, 44]. The small distance between

14

the molecules and graphene makes direct charge transfer much easier between

15

graphene and the molecules [1, 45]. Overall, graphene layer will slightly reduce

16

electric field enhancement, but could offer additional CM enhancement to improve

17

the total SERS properties. In the end, two enhancement mechanisms are combined to

18

give the overall SERS enhancement. For graphene-coated nanostructures, the stronger

19

chemical enhancement from graphene can potentially be used to improve the SERS

20

enhancement of Cu nanostructures.

AC C

EP

TE D

M AN U

SC

RI PT

3

16

RI PT

ACCEPTED MANUSCRIPT

SC

1

Fig. 7- Electric field distributions of 1× 3 distributed array particles on Cu

3

substrate (a-b) and SiO2 substrate (c-d) with diameter of 50 nm, nanogap of 15

4

nm. (a-c) and (b-d) represent the transverse and longitudinal section diagram,

5

respectively.

M AN U

2

To indentify the better SERS enhancement of G@CuNP/G@Cu than that of

7

G@CuNP/G/SiO2, we analyze the local electric field properties of these samples. Fig.

8

7 shows the electric field distributions of Cu nanoparticles deposited on Cu substrate

9

and SiO2 substrate. Obviously, the electric field of Cu nanoparticles on Cu substrate

10

is stronger than that on SiO2 substrate. The reason for this phenomenon is that local

11

surface plasmon will be formed between Cu nanoparticles and Cu substrate, which

12

couples with the local surface plasmon between Cu nanoparticles. For the SiO2

13

substrate case, the local surface plasmon between Cu nanoparticles is still formed,

14

while, the local surface plasmon can not be formed between nanoparticles and SiO2

15

substrate.

AC C

EP

TE D

6

17

SC

RI PT

ACCEPTED MANUSCRIPT

1

Fig. 8-The Raman spectra of the R6G molecules with concentrations from 10-9 M

3

on the G@CuNP/G@Cu and CuNP/Cu substrates before and after the surface

4

oxidation, respectively marked as G@CuNP/G@Cu (before oxidation),

5

O-G@CuNP/G@Cu (after oxidation), CuNP/Cu (before oxidation), O-CuNP/Cu

6

(after oxidation).

TE D

M AN U

2

More importantly, the stability of the G@CuNP/G@Cu substrate is improved,

8

benefiting from that graphene layer can be acted as the passivation layer to inhibit the

9

surface oxidation of the Cu nanoparticles and Cu substrate. The stability of the SERS

10

substrates based on G@CuNP/G@Cu and CuNP/Cu has been studied through a

11

thermal oxidation treatment. The thermal oxidation treatment was implemented by

12

expositing substrates to hot and humid air, with the temperature of 80 °C and the

13

humidity of 80%, for 5 days. R6G molecules with concentrations of 10−9 M were used

14

as the probe molecule. As shown in Fig. 8, the SERS intensity of R6G molecules on

15

G@CuNP/G@Cu substrate is nearly unchanged, indicating excellent chemical

16

stability of G@CuNPs. While for the oxidation treated CuNP/Cu substrate, the

AC C

EP

7

18

ACCEPTED MANUSCRIPT characteristic peaks of R6G molecules are absent. It has been known that CuNPs are

2

easily oxidized when they are exposed to air. The absence of the Raman signals can

3

be ascribed to the formation of copper oxides layer on the surface of Cu nanoparticles,

4

which has been confirmed by previous work [1]. Copper oxide layer acts as the

5

passivation layer to decrease the enhancement activity of samples. As oxygen gas and

6

moisture cannot permeate through the graphene layer, the graphene can effectively

7

protect metal from oxygen damage. The thermal stability comparison indicates that

8

the grown graphene shell on CuNPs and Cu substrate can more effectively suppress

9

degradation of the metallic nanostructures.

SC

M AN U

10

RI PT

1

4. Conclusions

In summary, we have proposed a novel, facilely synthesized and low-cost

12

graphene-Cu sandwich coupling system as the SERS substrate. The graphene covered

13

Cu nanoparticles are easily formed on the graphene@Cu foils by using the designed

14

growth process. The two enhancement mechanisms of the EM (Cu nanoparticles) and

15

CM (graphene) are combined to give the overall SERS enhancement of the fabricated

16

G@CuNP/G@Cu substrate. More importantly, the stability of the G@CuNP/G@Cu

17

substrate is improved, benefiting from that graphene layer can acts as the passivation

18

layer to inhibit the surface oxidation of the Cu nanoparticles and Cu substrate. The

19

SERS results show that the all-sided isolated structure is a promising substrate for

20

applications in chemical and biological detection because of its long-term stability and

21

superior SERS performance.

22

Acknowledgments

AC C

EP

TE D

11

19

ACCEPTED MANUSCRIPT The authors are grateful for financial support from the National Natural Science

2

Foundation of China (11474187, 11274204, 61205174 and 61307120), Specialized

3

research Fund for the Doctoral Program of Higher Education of China

4

(20133704120008), and Shandong Province Higher Educational Science and

5

Technology Program (J12LA07).

RI PT

1

6

SC

7

M AN U

8 9 10 11

15 16 17 18

EP

14

AC C

13

TE D

12

20

ACCEPTED MANUSCRIPT Reference

2

[1] Xu SC, Man BY, Jiang SZ, Wang J, Wei J, Xu S, et al. Graphene/Cu nanoparticle

3

hybrids fabricated by chemical vapor deposition as surface-enhanced Raman

4

scattering substrate for label-free detection of adenosine. ACS Appl. Mater. Inter.

5

2015; 7:10977-10987.

6

[2] Chen S, Li X, Zhao Y, Chang, LM, Qi, JY. Graphene oxide shell-isolated Ag

7

nanoparticles for surface-enhanced Raman scattering. Carbon 2015; 81:767-772.

8

[3] Li XH, Ren X, Zhang Y, Choy WC Wei B. An all-copper plasmonic sandwich

9

system obtained through directly depositing copper NPs on a CVD grown

M AN U

SC

RI PT

1

graphene/copper film and its application in SERS. Nanoscale 2015; 7(26):11291.

11

[4] Nie S, Emory S R. Probing single molecules and single nanoparticles by

12

surface-enhanced Raman scattering. Science 1997; 275(5303):1102-1106.

13

[5] Kiraly B, Yang S, Huang T J. Multifunctional porous silicon nanopillar arrays:

14

antireflection, superhydrophobicity, photoluminescence, and surface-enhanced Raman

15

scattering. Nanotechnology 2013; 24(24):245704.

16

[6] Zhang C, Man BY, Jiang SZ, Yang C, Liu M, Chen CS, et al. SERS detection of

17

low-concentration adenosine by silver nanoparticles on silicon nanoporous pyramid

18

arrays structure. App. Surf. Sci. 2015; 347:668-672.

19

[7] Chen J, Mårtensson T, Dick KA, Deppert K, Xu HQ, Samuelson L, et al.

20

Surface-enhanced Raman scattering of rhodamine 6G on nanowire arrays decorated

21

with gold nanoparticles. Nanotechnology 2008; 19(27):275712.

22

[8] Chen LM, Liu YN. Palladium crystals of various morphologies for SERS

AC C

EP

TE D

10

21

ACCEPTED MANUSCRIPT enhancement. CrystEngComm 2011; 13(21):6481-6487.

2

[9] Tan CL, Lee SK, Lee YT. Bi-SERS sensing and enhancement by Au-Ag bimetallic

3

non-alloyed nanoparticles on amorphous and crystalline silicon substrate. Opt.

4

Express 2015; 23(5): 6254-6263.

5

[10] Hwang HJ, Jung CH, Choi JH, Lee SY, Park DJ, Ha KS. Surface enhanced

6

Raman scattering by gold nanoparticle-decorated reduced graphene oxide on

7

ITO-coated glass. Sci. Adv. Mater. 2014; 6(11):2566-2571.

8

[11] Yang H, Ni SQ, Jiang X, Jiang W, Zhan JH. In situ fabrication of

9

single-crystalline porous ZnO nanoplates on zinc foil to support silver nanoparticles

M AN U

SC

RI PT

1

as a stable SERS substrate. CrystEngComm 2012; 14(18):6023-6028.

11

[12] Cheng L, Shao M, Zhang M. An ultrasensitive method to detect dopamine from

12

single mouse brain cell: surface-enhanced Raman scattering on Ag nanoparticles from

13

beta-silver vanadate and copper. S Sci. Adv. Mater. 2010; 2(3):386-389.

14

[13] Li X, Li J, Zhou X, Ma Y, Zheng Z, Duan X, et al. Silver nanoparticles protected

15

by monolayer graphene as a stabilized substrate for surface enhanced Raman

16

spectroscopy. Carbon 2014; 66:713-719.

17

[14] Fan M, Andrade GF, Brolo AG. A review on the fabrication of substrates for

18

surface enhanced Raman spectroscopy and their applications in analytical chemistry.

19

Anal. Chim. Acta 2011; 693:7-25.

20

[15] Banholzer MJ, Millstone JE, Qin L, Mirkin CA. Rationally designed

21

nanostructures for surface-enhanced Raman spectroscopy. Chem. Soc. Rev. 2008;

22

37(5):885-897.

AC C

EP

TE D

10

22

ACCEPTED MANUSCRIPT [16] Zheng J, Li X, Gu R, Lu T. Comparison of the surface properties of the

2

assembled silver nanoparticle electrode and roughened silver electrode. J. Phys. Chem.

3

B 2002; 106(5):1019-1023.

4

[17] Li JF, Huang YF, Ding Y, Yang ZL, Li SB, Zhou XS, et al. Shell-isolated

5

nanoparticle-enhanced Raman spectroscopy. Nature 2010; 464(7287):392-395.

6

[18] Zheng C, Liang L, Xu S, Zhang H, Hu C, Bi L, et al. The use of Au@ SiO2

7

shell-isolated nanoparticle-enhanced Raman spectroscopy for human breast cancer

8

detection. Anal. and Bioanal. Chem 2014; 406(22): 5425-5432.

9

[19] Vaidya S, Patra A, Ganguli AK. Optical properties of Ag@ TiO2 and CdS@ TiO2

M AN U

SC

RI PT

1

core-shell nanostructures. Sci. Adv. Mater. 2012; 4(5-6):631-636.

11

[20] Zhang C, Man B Y, Yang C, Jiang SZ, Liu M, Chen CS, et al. Facile synthesis of

12

graphene on dielectric surfaces using a two-temperature reactor CVD system.

13

Nanotechnology 2013; 24(39):395603.

14

[21] Xu W, Ling X, Xiao J, Dresselhaus M, Kong J, Xu H, et al. Surface enhanced

15

Raman spectroscopy on a flat graphene surface. Proceedings of the National Academy

16

of Sciences 2012; 109(24): 9281-9286.

17

[22] Zhang C, Man B Y, Yang C, Jiang SZ, Liu M, Chen CS, et al. Direct formation of

18

graphene-carbon nanotubes hybrid on SiO2 substrate via chemical vapor deposition.

19

Sci. Adv. Mater. 2014; 6(2):399-404.

20

[23] Yang C, Xu Y, Zhang C, Sun Z, Chen C, Li X, et al. Facile synthesis 3D flexible

21

core-shell graphene/glass fiber via chemical vapor deposition. Nanoscale Res. Lett.

22

2014; 9(1):394.

AC C

EP

TE D

10

23

ACCEPTED MANUSCRIPT [24] Liu Y, Hu Y, Zhang J. Few-layer graphene-encapsulated metal nanoparticles for

2

surface-enhanced Raman spectroscopy. J. Phys. Chem. C 2014; 118(17):8993-8998.

3

[25] Xu W, Mao N, Zhang J. Graphene: A Platform for Surface‐Enhanced Raman

4

Spectroscopy. Small 2013; 9(8): 1206-1224.

5

[26] Xie L, Ling X, Fang Y, Zhang J, Liu Z. Graphene as a substrate to suppress

6

fluorescence in resonance Raman spectroscopy. J. Am. Chem. Soc. 2009;

7

131(29):9890-9891.

8

[27] Ling X, Zhang J. Interference phenomenon in graphene-enhanced Raman

9

scattering. J. Phys. Chem. C 2011; 115(6):2835-2840.

M AN U

SC

RI PT

1

10

[28] Zhang Z, Xu F, Yang W, Guo M, Wang X, Zhang B, et al. A facile one-pot

11

method

12

surface-enhanced Raman scattering. Chem. Commun. 2011; 47(22):6440-6442.

13

[29] He S, Liu K K, Su S, Yan J, Mao X, Wang D, et al. Graphene-based

14

high-efficiency surface-enhanced Raman scattering-active platform for sensitive and

15

multiplex DNA detection. Anal. Chem. 2012; 84(10):4622-4627.

16

[30] Johnson PB, Christy RW. Optical constants of the noble metals. Phys. Rev. B

17

1972; 6(12):4370.

18

[31] Schedin F, Lidorikis E, Lombardo A, Kravets VG, Geim AK, Grigorenko AN, et

19

al. Surface-enhanced Raman spectroscopy of graphene. ACS Nano 2010;

20

4(10):5617-5626.

21

[32] Hao Q, Wang B, Bossard JA, Kiraly B, Zeng Y, Chiang I, et al. Surface-enhanced

22

Raman scattering study on graphene-coated metallic nanostructure substrates. J. Phys.

high-quality

Ag-graphene

composite

nanosheets

for

efficient

AC C

EP

TE D

to

24

ACCEPTED MANUSCRIPT Chem. C 2012; 116(13):7249-7254.

2

[33] Ferrari AC, Meyer JC, Scardaci V, Casiraghi C, Lazzeri M, Mauri F, et al. Raman

3

spectrum of graphene and graphene layers. Phys. Rev. Lett. 2006; 97(18):187401.

4

[34] Persson BNJ, Zhao K, Zhang Z. Chemical contribution to surface-enhanced

5

Raman scattering. Phys. Rev. Lett. 2006; 96(20):207401.

6

[35] Otto A, Mrozek I, Grabhorn H, Akemann WJ. Surface-enhanced Raman

7

scattering. J. Phys: Condens. Matter 1992; 4(5):1143.

8

[36] Novoselov KS, Geim AK, Morozov SV, Jiang D, Zhang Y, Dubonos SV, et al.

9

Electric field effect in atomically thin carbon films. Science 2004; 306(5696):666-669.

10

[37] Wang P, Liang O, Zhang W, Schroeder T, Xie YH. Ultra-sensitive

11

graphene-plasmonic hybrid platform for label-free detection. Adv. Mater. 2013;

12

25(35):4918-4924.

13

[38] Michaels AM, Jiang J, Brus L. Ag nanocrystal junctions as the site for

14

surface-enhanced Raman scattering of single rhodamine 6G molecules. J. Phys. Chem.

15

B 2000; 104(50):11965-11971.

16

[39] Zhao J, Jensen L, Sung J, Zou S, Schatz GC, Van Duyne RP. Interaction of

17

plasmon and molecular resonances for rhodamine 6G adsorbed on silver nanoparticles.

18

J. Am. Chem. Soc. 2007; 129(24):7647-7656.

19

[40] Ferrari AC, Meyer JC, Scardaci V, Casiraghi C, Lazzeri M, Mauri F,et al. Raman

20

spectrum of graphene and graphene layers. Phys. Rev. Lett. 2006; 97(18):187401.

21

[41] Moskovits M. Surface-enhanced spectroscopy. Rev. Mod. Phys. 1985; 57(3):783.

22

[42] Kneipp K, Kneipp H, Itzkan I, Dasari RR, Feld MS. Ultrasensitive chemical

AC C

EP

TE D

M AN U

SC

RI PT

1

25

ACCEPTED MANUSCRIPT analysis by Raman spectroscopy. Chem. Rev. 1999; 99(10):2957-2976.

2

[43] Hong G, Wu Q H, Ren J, Wang C, Zhang W, Lee ST. Recent progress in organic

3

molecule/graphene interfaces. Nano Today 2013; 8(4):388-402.

4

[44] Thrall E S, Crowther A C, Yu Z, Brus LE. R6G on graphene: high Raman

5

detection sensitivity, yet decreased Raman cross-section. Nano Lett 2012; 12(3):

6

571-1577.

7

[45] Singh AK, Khan SA, Fan Z, Demeritte T, Senapati D, Kanchanapally R, et al.

8

Development of a long-range surface-enhanced Raman spectroscopy ruler. J J. Am.

9

Chem. Soc. 2012; 134(20):8662-8669.

M AN U

SC

RI PT

1

AC C

EP

TE D

10

26