Enhancing SERS Sensitivity in N‐Graphene Hydrangea by Synergistic Charge‐Transfer and Excitation Light Absorption

Surface‐enhanced Raman scattering (SERS) is a versatile spectroscopic technique, which plays a crucial role in enhancing analytical sensitivity, investigating interfacial reaction mechanisms, enabling biosensing, and fostering efficient catalysis. Currently, the common SERS substrates are primarily metal nanostructures, which entail high manufacturing costs, complex processes, and the metal surface undergo change over time and with environmental conditions. These issues limit the development of SERS technology. In this work, a nitrogen‐doped graphene (N‐graphene) hydrangea was synthesized on a silicon (Si) substrate using plasma‐assisted chemical vapor deposition (PACVD), forming an N‐graphene hydrangea/Si hybrid structure as a SERS substrate. This substrate offers the advantages of high stability, ultra‐sensitivity, and reusability. The three‐dimensional nano‐cavity structure of graphene can increase the interaction between light and graphene, resulting in an increased localized electric field. Combining theoretical simulation analysis, the introduction of nitrogen (N) elements adjusts the Fermi level of graphene, promoting efficient charge transfer. In practical scenarios, Di(2‐ethylhexyl) phthalate (DEHP), a commonly used plasticizer, has raised concerns due to its potential as an endocrine disruptor and carcinogen. The as‐prepared SERS substrate achieves a remarkable detection limit of as low as 10−8 m for DEHP, providing significant support for environmental conservation and human health.


Surface-enhanced
Raman scattering (SERS), a spectral analytical technique, plays a crucial role in surface analysis, enhancing the detection and characterization capability of molecules.Boasts a history of over 20 years.Its origins trace back to the 1970s when researchers initially noted a remarkable amplification in Raman scattered signals when molecules or nanostructures were present on metal surfaces. [1,2][5] It enables the detection of molecules at low concentrations, characterization of organic molecular structures, and precise analysis in biological studies, making it highly valuable for both scientific research and industrial applications. [6,7]Following this discovery, the researchers commenced theoretical investigations to elucidate the mechanism underlying the SERS effect.These investigations encompassed electromagnetic theory, the local electric field enhancement effect, and the chemical enhancement effect. [8,9]The enhancement effect in SERS primarily involves electromagnetic enhancement (EM) and chemical enhancement (CM). [8,9]EM results from the localized electric field Surface-enhanced Raman scattering (SERS) is a versatile spectroscopic technique, which plays a crucial role in enhancing analytical sensitivity, investigating interfacial reaction mechanisms, enabling biosensing, and fostering efficient catalysis.Currently, the common SERS substrates are primarily metal nanostructures, which entail high manufacturing costs, complex processes, and the metal surface undergo change over time and with environmental conditions.These issues limit the development of SERS technology.In this work, a nitrogendoped graphene (N-graphene) hydrangea was synthesized on a silicon (Si) substrate using plasma-assisted chemical vapor deposition (PACVD), forming an N-graphene hydrangea/Si hybrid structure as a SERS substrate.This substrate offers the advantages of high stability, ultra-sensitivity, and reusability.The three-dimensional nano-cavity structure of graphene can increase the interaction between light and graphene, resulting in an increased localized electric field.Combining theoretical simulation analysis, the introduction of nitrogen (N) elements adjusts the Fermi level of graphene, promoting efficient charge transfer.In practical scenarios, Di(2-ethylhexyl) phthalate (DEHP), a commonly used plasticizer, has raised concerns due to its potential as an endocrine disruptor and carcinogen.The as-prepared SERS substrate achieves a remarkable detection limit of as low as 10 À8 M for DEHP, providing significant support for environmental conservation and human health.[20][21] The common SERS substrates employing precious metal nanoparticles pose challenges in precise control regarding shape and size, limiting their practical application due to high cost. [22]Achieving rapid, accurate, and high-throughput SERS detection faces challenges concerning substrate design, fabrication repeatability, and stability issues. [23,24]SERS has emerged as a widely adopted spectral technique for detecting and analyzing trace molecules, encompassing biomolecules, organic compounds, and inorganic substances.Recent years have witnessed significant advancements in SERS technology propelled by the rapid progress in nanotechnology. [6,7]The meticulous preparation of nanomaterials with precise shape and structure has enabled more accurate SERS enhancement, thereby amplifying the sensitivity of analytical measurements.In recent years, two-dimensional (2D) graphene and its derivatives have gained widespread attention as substrate materials for SERS. [25,26]However, the low optical absorption of 2D-graphene, with only 2.3%, significantly limits its detection performance as a SERS substrate. [27,28]In contrast to traditional 2D-graphene, three-dimensional (3D) graphene possesses a highly porous structure and a larger specific surface area, providing abundant adsorption sites.Its unique nanocage structure enables stronger light absorption, endowing it with superior optical performance. [29,30][33] N atoms also act as reaction centers to adsorb molecules or ions, facilitating key reactions such as adsorption. [34,35]Forming the nitrogen-doped graphene(N-graphene) hydrangea structure, it has a larger surface area, which means it can provide more active sites for molecular adsorption and reactions.This multidimensional structure increases the interaction between molecules and materials, helping to enhance adsorption and reaction.At the same time, it has better stability and is less susceptible to the influence of external environmental factors.This makes it more reliable during long-term use and storage.This choice is innovative in itself, offering the potential for highly sensitive SERS properties.This increase in complex structure and surface area increased the intensity of SERS effect and increased the interaction between molecules and substrate, thus enhancing SERS signal.These properties give 3D-graphene distinctive advantages in SERS applications.
In this study, we employed plasma-assisted chemical vapor deposition (PACVD) to synthesize an N-graphene hydrangea/ Si hybrid structure for high-performance SERS detection.The as-prepared SERS boasts outstanding properties, including excellent adsorption capacity, a large specific surface area, porosity, and exceptional optoelectronic properties.N-graphene hydrangea features a complex structure and a larger surface area, allowing for the localization of electromagnetic field enhancement within the nanocage structure.This localized electric field enhancement concentrates the electromagnetic energy of incident light near the surface of the nanocage structure, [36] thereby increasing the interaction between light and graphene hydrangea. [37,38]ese critical "hot spot" regions, usually located between the structures of N-graphene hydrangeas, have important SERS properties.These sharp nanotips and tiny gaps or cavities not only enhance the interaction between light and matter, [39] they also provide more active sites, making it easier for molecules to adsorb to the surface.This further improves the sensitivity and reliability of the SERS effect.Furthermore, the introduction of N atoms can adjust the Fermi level position of graphene hydrangea, enhancing its conductivity and improving its performance in electronic and charge transfer processes. [31,33,40]Due to the distinct electron affinity and electron-donating properties of N atoms, they serve as reaction centers for the adsorption of molecules or ions, significantly enhancing charge transfer and adsorption processes. [34,35]Through first-principles density functional theory (DFT) simulation and scanning Kelvin probe microscopy (SKPM) analysis, demonstrating that N-graphene hydrangea exhibits a stronger electron-donating ability, enhancing the interaction between the probe molecules and the substrate, thereby improving charge transfer in a hybrid structure.Experimental results indicate that the as-prepared SERS substrate has the lowest detection limits for probe molecules such as erythrosine B (EB), crystal violet (CRV), and rhodamine 6G (R6G) are 10 À11 , 10 À8 , and 10 À11 M, respectively.In practical applications, this substrate can detect DEHP at a minimum concentration of 10 À8 M, reaching the level of micromolar, and successfully detecting DEHP in perfume samples.Therefore, the N-graphene hydrangea/Si hybrid structure demonstrated high sensitivity, long-term stability, high reproducibility, and reusability advantages.As a SERS detection substrate, it holds great potential for applications in the detection of biomolecules and environmental hormones.

Characterization of N-Graphene Hydrangea/Si Hybrid Structures
The surface scanning electron microscopy (SEM) image of N-graphene hydrangea/Si hybrid structure was depicted in Figure 1a, revealing the cluster structure of graphene, with individual hydrangea clusters having an approximate diameter of 1 μm. Figure 1b illustrates that N-graphene hydrangea/Si possesses around 90% porosity, offering a larger specific surface area and more adsorption sites. [41]The atomic force microscopy (AFM) image is shown in Figure 1c, confirming the vertical arrangement of graphene on the Si substrate.The height of N-graphene hydrangea was demonstrated in the supplementary information in Figure S1, Supporting Information.Figure 1d presents a 3D model of a single N-graphene hydrangea.The finite-different time-domain (FDTD) simulation was explored the interaction between incident light and N-graphene hydrangea.Figure 1e displays the normalized local electric field distribution of N-graphene hydrangea.The simulated 3D-view of the normalized power dissipation density distribution in N-graphene hydrangea was shown in Figure 1f, revealing that incident light absorption occurs within the graphene structure, with the maximum absorption location at the edges of N-graphene hydrangea.A comparison of the Raman spectra between N-graphene hydrangea and pristine graphene hydrangea is depicted in Figure 1g.Pristine graphene exhibits three characteristic peaks: the D peak (≈1350 cm À1 ), the G peak (≈1580 cm À1 ), and the 2D peak (≈2700 cm À1 ).The introduction of N atoms affects the crystallinity and orderliness of graphene, inducing interlayer coupling between N atoms and graphene, leading to a redshift of the G peak and a blueshift of the 2D peak in N-graphene hydrangea. [42]Figure 1h,i present high-resolution XPS spectra of C-1s and N-1s, respectively, revealing the chemical states of N-graphene hydrangea.In Figure 1h, the C-1s spectrum can be fitted with three peaks (≈284.0,≈284.7, and ≈289.1 eV), corresponding to C=C, C=N, and C-N, respectively. [43,44]The C-1s spectrum of pristine graphene hydrangea/Si without N doping is provided in the Figure S2, Supporting Information.As shown in Figure 1i, the N-1s spectrum consists of three fitted peaks at ≈398.4,≈401.3, and ≈403.2 eV, indicating the presence of pyridinic N, pyrrolic N, and graphitic N, [45,46] confirming that the successful incorporation of N atoms into the graphene lattice, forming N-doped graphene.Further comparison of N-graphene hydrangea and pristine graphene hydrangea is shown in Figure S3,S4, Supporting Information.N-graphene hydrangea possesses a higher porosity, resulting in a larger surface area and increased interface for interaction with incident light.Moreover, its unique nanocavity structure and cavity resonance effect can enhance light absorption. [47]The optical absorption spectrum of the N-graphene hydrangea/Si system is depicted in Figure S5, Supporting Information.Furthermore, the introduction of N atoms regulates the Fermi level of graphene, facilitating charge transfer. [48]Therefore, N-graphene hydrangea/Si exhibits high sensitivity as a SERS substrate.

The Charge Transfer between Probe Molecules and the Graphene System was Investigated by Simulation Method
To investigate the photoinduced charge transfer (PICT) of N-graphene hydrangea/Si heterojunctions, the density of electron states (DOS) results by comparing N-graphene hydrangea/Si with graphene hydrangea/Si using the DFT simulation, as shown in Figure S6, Supporting Information.Compared to the graphene hydrangea/Si, N-graphene hydrangea/Si exhibits a greater number of electronic states near the Fermi level, primarily contributed by the C p orbitals and N atoms.The DOS analysis indicates that the N-graphene hydrangea/Si system enhances the electron-transfer ability, effectively improving the PICT resonance and SERS effect.To further elucidate why N-graphene hydrangea/Si substrates exhibit higher SERS enhancement, the electrostatic potentials of the two systems in the Z-direction are discussed, as demonstrated in Figure 2a.hydrangea/Si (5.430 eV) is larger than that of graphene hydrangea/Si (5.401 eV).Additionally, Figure 2b reveals that the Fermi levels of graphene hydrangea/Si and N-graphene hydrangea/Si both lie above the lowest unoccupied molecular orbital (LUMO) of the DEHP molecule (À2.396 eV), with the Fermi level of the N-graphene hydrangea/Si being closer to the LUMO of the adsorbed molecule.Consequently, this effectively promotes charge transfer between the DEHP molecule and N-graphene hydrangea/Si, resulting in improved SERS performance. [49]or further investigation into the charge transfer between the DEHP molecule and substrates, we constructed DEHP and graphene hydrangea/Si as well as DEHP and N-graphene hydrangea/Si structures, as illustrated in Figure 2c,f.Subsequently, the charge density difference and worse charge between the substrates and adsorbed molecules were calculated, as displayed in Figure 2d,e,g,h.The results indicate that 0.498 e À is transferred from graphene hydrangea/Si to DEHP molecule, whereas 0.589 e À is transferred from N-graphene hydrangea/Si to DEHP molecule, highlighting the stronger electron-donation ability of the N-doped system.

Research on SERS Detection Mechanism of N-Graphene Hydrangea/Si
To demonstrate the enhanced detection performance of N-graphene hydrangea/Si as a SERS substrate through charge transfer-based chemical enhancement, we employed SKPM and conductive atomic force microscopy (CAFM) to investigate the carrier transfer between the probe molecules and the substrate.The surface potential distribution of N-graphene hydrangea/Si before and after DEHP adsorption was measured using SKPM, as shown in Figure 3a,b.Given charge transfer characteristics, surface potential distribution closely correlates with the distribution of carriers (charges) on the graphene surface.The significant increase in surface potential after DEHP adsorption indicates efficient charge transfer between DEHP molecules and N-graphene hydrangea/Si, resulting in surface potential changes reflecting the distribution of probe carriers within graphene.Furthermore, compared the current of DEHP/N-graphene hydrangea/Si under dark and illuminated conditions using CAFM, as illustrated in Figure 3c,d.The study results demonstrate that the photocurrent generated by graphene is significantly higher under illumination compared to dark conditions.Spatial current distribution curves under dark and illuminated conditions are depicted in Figure 3e,f, respectively, illustrating the variation in current values.This confirms the presence of photo-induced carrier transfer between the probe molecule DEHP and the N-graphene hydrangea/Si, thus validating the significant contribution of the chemical enhancement mechanism to the enhanced SERS detection performance.

N-Graphene Hydrangea/Si Heterogeneous Structure was used to Build a Sensitive, Efficient, and Stable SERS Platform
The as-prepared N-graphene hydrangea/Si substrate was utilized as a SERS substrate to detect the probe solutions of fluorescent organic molecules and its SERS detection performance was investigated.Figure 4a-c present the detection results of EB, CRV, and R6G at the same concentration on three different SERS substrates (N-graphene hydrangea/Si, graphene hydrangea/Si, and Si). Figure S7, Supporting Information summarizes the measurement sensitivity plots, representing the peak intensities of R6G and CRV molecules, indicating that the N-graphene hydrangea/Si exhibits excellent light absorption capabilities and remarkable molecular adsorption efficiency among the three SERS substrates.
In this work, EB was selected as the target substance, and the peak intensity of the characteristic peak of EB at 1344 cm À1 was obtained for both surface-enhanced Raman spectroscopy and conventional Raman spectroscopy.By substituting these values into the calculation formula, [50] an enhancement factor (EF) of 3.52 Â 10 6 was determined for the N-graphene hydrangea/Si substrate.Figure 4d-f show the Raman spectra of EB, CRV, and R6G at different concentrations on the N-graphene hydrangea/Si substrate, demonstrating that the intensity of the characteristic peaks of different probe molecules is positively correlated with their concentrations.The detection limits of EB, CRV, and R6G were found to be as low as 10 À11 , 10 À8 , and 10 À11 M, respectively.Additionally, linear fitting was performed on the relationship between the peak intensities and the concentrations, as shown in Figure 4g-i.The linear correlation coefficients (R 2 ) for EB at 1344 and 1605 cm À1 were 0.99351 and 0.98992, respectively.For CRV, the R 2 values at 1371 and 1621 cm À1 were 0.99514 and 0.99083, respectively.For R6G, the R 2 values at 1360 and 1647 cm À1 were 0.98944 and 0.98943, respectively.Therefore, N-graphene hydrangea/Si holds great potential as a SERS detection substrate.
Uniformity, stability, and reproducibility are crucial criteria for assessing the detection performance of SERS substrates. [51,52]herefore, Raman spectra were randomly collected from 30 points on the N-graphene hydrangea/Si substrate, as depicted in Figure 5a.The Raman intensity distributions of the D peak (Figure 5b), G peak (Figure 5c), and 2D peak (Figure S8, Supporting Information) were recorded from 900 randomly selected points on the substrate surface.The relative standard deviations (RSD) were calculated to be 3.63%, 3.44%, and 4.20%, respectively.Similarly, to demonstrate the detection performance of the substrate for DEHP, Raman spectra were randomly collected from 30 points on the N-graphene hydrangea/Si substrate with 10 À4 M DEHP, as illustrated in Figure 5d.The Raman intensity distributions of DEHP at 1039, 1349, and 1600 cm À1 were recorded from 900 randomly selected points on the substrate surface.The calculated RSD were 4.01%, 3.66%, and 3.30%, respectively, as shown in Figure 5e-f and S9, Supporting Information.These subtle differences highlight the excellent uniformity and stability of N-graphene hydrangea/ Si as a SERS detection substrate.
Raman spectra were randomly collected from 30 points on the N-graphene hydrangea/Si substrate with 10 À4 M EB (Figure 6a), CRV (Figure 6b), and R6G (Figure 6c) to confirm the excellent detection performance of the substrate for probe molecules.The SERS spectra of EB, CRV, and R6G detection on the N-graphene hydrangea/Si substrate after five ethanol washes are shown in Figure 6e-f.This also demonstrates the substrate's good reusability.The results indicate that N-graphene hydrangea/Si substrate is a reliable SERS platform with good uniformity, stability, and reproducibility.

Practical Application of N-Graphene Hydrangea/Si as SERS Substrate in DEHP Detection
In recent years, the use of phthalates in perfumes has garnered attention, as these chemicals are sometimes intentionally added to enhance their stability and longevity. [53,54]57] However, DEHP has demonstrated potential health risks, including potential disruption of the endocrine system, impacts on reproductive development, and effects on thyroid function in the human body. [58,59][67][68] Therefore, we have explored the practical application of a novel N-graphene hydrangea/Si substrate for SERS detection of DEHP in perfumes.
From Figure 7a, it is evident that the characteristic Raman peaks of DEHP are located at 1039, 1155, 1349, 1445, and 1599 cm À1 .The N-graphene hydrangea/Si substrate can detect DEHP in perfumes diluted to a concentration of 10 À4 , with a detection limit as low as 10 À8 M for pure DEHP, as shown in Figure S10a, Supporting Information.Linear fitting of the intensities of the two characteristic peaks of DEHP in perfumes is presented in Figure 7b, with R 2 values reaching approximately 0.9672 and 0.9805.Similarly, linear fitting of the intensities of the two characteristic peaks of pure DEHP is shown in Figure S10b, Supporting Information, with R 2 values of approximately 0.9957 and 0.9935, highlighting the quantitative detection capability of the substrate for DEHP molecules.To validate the long-term stability of the as-prepared SERS substrate, storage monitoring was conducted for up to 30 days, as shown in Figure 7c and S10c, Supporting Information where the intensity of the characteristic peaks of DEHP at 1445 and 1039 cm À1 did not significantly change over time.Additionally, reproducibility experiments were conducted on the substrate, and Figure 7d,e show the 3D-AFM images and the contact angles of N-graphene hydrangea/Si before and after washing, indicating no significant changes in surface morphology.The SERS spectra of DEHP in  perfumes and pure DEHP solution on the N-graphene hydrangea/Si substrate after washing are shown in Figure 7f and S10d, Supporting Information, respectively, indicating that the intensities of the characteristic peaks of DEHP at 1349, 1445, and 1599 cm À1 remain almost unchanged, demonstrating the excellent stability and reusability of the substrate.

Conclusion
In summary, high-performance SERS substrates based on N-graphene hydrangea/Si were demonstrated using a simple PACVD method to detect multiple analytes simultaneously.The excellent sensitivity and stability of the as-prepared SERS detection substrate were validated using EB, CRV, and R6G as probe molecules, with detection limits of 10 À11 , 10 À8 , and 10 À11 M, respectively.The hydrangea structure provides high porosity and larger surface area, while its nanocavity structure enhances the interaction between incident light and graphene, resulting in enhanced local field effects, improved SERS enhancement factors, and increased sensitivity.Furthermore, the introduction of N into graphene helps regulate the Fermi level, facilitating more efficient charge transfer processes and further enhancing the SERS performance.Overall, the N-graphene hydrangea/Si hybrid structure significantly amplifies the SERS signal through the synergistic effects of charge transfer and localized surface plasmon resonance.This enhancement effect, resulting from the combined action of chemical and electromagnetic mechanisms, opens up new possibilities for the application of SERS technology.This research provides a highly sensitive, stable, and reusable SERS substrate, and the N-graphene hydrangea/Si substrate holds great potential for applications in the detection of biological molecules and environmental hormones.

Experimental Section
N-Graphene Hydrangea/Si Substrate Fabrication: The Si substrate with dimensions of 1 cm Â 1 cm was placed in a quartz tube of a PACVD system (Equipment model BTF-1200C-II-AS-PACVD, purchased from Anhui Beyike Equipment Technology Co., Ltd.).The tube was initially evacuated to a pressure of around 5 Pa, followed by the introduction of a mixed gas of 1 sccm hydrogen (H 2 ) and 10 sccm argon (Ar).The temperature was then raised to 550 °C.Subsequently, the H 2 and Ar gases were turned off, and 15 sccm methane (CH 4 ) was introduced as the growth gas.The plasma source was ignited by setting the radio frequency power to 200 W.After 60 min of growth, the CH 4 flow was stopped, and the plasma source was turned off.The temperature was further increased to 750 °C, and 15 sccm of CH 4 and 1 sccm of ammonia (NH 3 ) were introduced.The plasma source was ignited again with a radio frequency power of 250 W, and the growth continued for another 30 min.At the end of the growth time, CH 4 and NH 3 were turned off, and 10 sccm of Ar was introduced to bring the pressure in the tube back to atmospheric pressure.The sample was then allowed to cool down before being removed.
Preparation of Probe Molecular Materials: EB, CRV, R6G, and DEHP were purchased from Aladdin Reagent (Shanghai, China) as dye probe molecules and dissolved in ethanol.The detection concentration range for EB and R6G was from 10 À4 to 10 À11 M, for CRV was from 10 À4 to 10 À8 M, and for DEHP was from 10 À2 to 10 À8 M. In addition, DEHP in perfumes was also tested, and even after diluting the perfume with ethanol by a factor of 10 4 , it could still be detected by the substrate.
Characterization: The surface morphology and height of the N-graphene hydrangea/Si substrate were measured using an AFM (Oxford Instruments Cypher S), and a SEM (HITAGHT S-3400N).These measurements provided high-resolution surface morphology images and accurate height information of the substrate.Additionally, the surface properties and wetting performance of the N-graphene hydrangea were evaluated by measuring the water contact angle on the material surface using a contact angle meter (JCA-1).To investigate the potential distribution and current response of the N-graphene hydrangea/Si, a SKPM and Conductive AFM (C-AFM) were employed, revealing the electronic behavior and conductivity of the material.Furthermore, to assess the structural quality of the N-graphene hydrangea, absorption spectra were collected using ultraviolet-visible-near-infrared spectroscopy (Cary 5000), and Raman scattering analysis was performed using a laser with a wavelength of 532 nm (Jobin Yvon HR800).Through these advanced characterization techniques, we were able to gain in-depth insights into and evaluate the physical and chemical properties of the N-graphene hydrangea/Si material, providing important references for its applications and optimization.
SERS Measurements: First, take 5 μL of the probe molecule solution and drop it onto the SERS substrate.Then, allow the solution to naturally dry on the surface of the substrate to ensure uniform distribution of the sample.During the SERS detection process, select a laser with a wavelength of 532 nm as the excitation source.The laser spot diameter is set to 12.5 μm to ensure sufficient illumination on the sample surface.To maintain the accuracy and reliability of the data, set a constant laser power of 2.5 mW and a data acquisition time of 10 s.This ensures sufficient collection of SERS signals from the sample surface to obtain accurate spectral information.Through these steps, enhanced spectral signals of the probe molecules can be obtained.These data will help us evaluate the performance and potential applications of the N-graphene hydrangea/Si material.
Computational Methods: All calculations were carried out using the plane-wave projector augmented-wave (PAW) method, [69] applying the Perdew-Burke-Ernzerhof (PBE) exchange-correlation functional, [70] as implemented in the Vienna ab initio simulation package (VASP) code. [71] plane-wave cutoff energy of 400 eV was used.A vacuum space of 15 Å was added to the non-periodic direction to avoid interaction between periodic images for slab structures.Brillouin-zone integration was performed on gamma-centered symmetry-reduced meshes.The Brillouin zone is sampled with Gama-centered 2 Â 2 Â 1 k-point meshes for the optimization of slab structures. [72]The structures were relaxed with the implemented conjugate-gradient algorithm, until the convergence tolerance of Hellmann-Feynman forces and energy on each atom were less than 0.1 eV Å À1 and 10 À5 eV atom À1 , respectively.What's more, the DFT-D3 method was included to improve the description of the long-range weak van der Waals (vdW) interaction for all DFT calculations.
Simulation Method: The wavelength-dependent response of the N-graphene hydrangea structure to linearly polarized light and related electric field distributions and power loss density were simulated under periodic boundary conditions, using the FDTD method.The simulation domain in the light propagation direction (along the hydrangea axis) was terminated by perfect match layers.Complex permittivity for N-graphene hydrangea was interpolated from bulk dielectric values reported by Johnson and Christy, and modified by considering the size effects of N-graphene hydrangea.The complex relative permittivity of graphene can be generally expressed as a function of wavelength ε(λ) = 5.5 þ i λ/246d, where d is the thickness of graphene which is about 3.8 Å.

Figure 1 .
Figure 1.a) SEM image of N-graphene hydrangea/Si surface.b) Porosity image of N-graphene hydrangea/Si.c) 3D-AFM image of N-graphene hydrangea/Si.d) 3D-model diagram of N-graphene hydrangea.e) Normalized electric field distribution of N-graphene hydrangea/Si.f ) Spatial distribution of normalized power loss density of N-graphene hydrangea/Si.g) Comparison of Raman spectra of N-graphene hydrangea/Si and graphene hydrangea/Si.h) High-resolution C-1s XPS spectra of N-graphene hydrangea/Si.i) High-resolution N-1s XPS spectrum of N-graphene hydrangea/Si.

Figure 2 .
Figure 2. a) Electrostatic potential of graphene hydrangea/Si structures and N-graphene hydrangea/Si structures.b) Ground state charge transfer between DEHP and graphene hydrangea/Si structures and N-graphene hydrangea/Si structures.c) Front view of DEHP and graphene hydrangea/Si model.d) Charge distribution diagram of DEHP and graphene hydrangea/Si.e) Side view of charge distribution of DEHP and graphene hydrangea/Si.f ) Front view of DEHP and N-graphene hydrangea/Si models.g) Charge distribution diagram of DEHP and N-graphene hydrangea/Si.h) Side view of charge distribution of DEHP and N-graphene hydrangea/Si.

Figure 3 .
Figure 3. a) Surface potential distribution of N-graphene hydrangea/Si.b) Surface potential distribution N-graphene hydrangea/Si after DEHP adsorption.c,d) Current graphs on N-graphene hydrangea/Si measured using DEHP under dark and light conditions, respectively.e,f ) The current distribution corresponding to (c,d), respectively.

Figure 4 .
Figure 4. SERS spectra of a) EB, b) CRV, and c) R6G as probes on N-graphene hydrangea/Si, Graphene hydrangea/Si and Si.Detection sensitivity of N-graphene hydrangea/Si for d) EB, e) CRV, and f ) R6G, respectively.Logarithmic curves of different characteristic peaks of g) EB, h) CRV, and i) R6G with concentration in Raman spectra.

Figure 5 .
Figure 5. a) Raman spectra of 30 points in N-graphene hydrangea/Si are randomly collected.Raman intensity distribution of peaks b) D and c) G in N-graphene hydrangea/Si.d) Random sampling of 30-point Raman spectra on a 10 À4 M DEHP N-graphene hydrangea/Si substrate.Raman intensity distribution of characteristic peaks of DEHP at e) 1039 cm À1 and f ) 1349 cm À1 .

Figure 6 .
Figure 6.Raman spectra of 30 points were randomly collected on N-graphene hydrangea/Si substrates of 10 À4 M a) EB, b) CRV, and c) R6G.SERS spectra of d) EB, e) CRV, and f ) R6G were detected by N-graphene hydrangea/Si after cleaning.

Figure 7 .
Figure 7. a) Detection sensitivity of N-graphene hydrangea/Si to DEHP in perfumes.b) Logarithmic curves of different characteristic peaks of DEHP in perfume in Raman spectrum with concentration variation.c) Long-term stability of the DEHP spectra recorded on the N-graphene hydrangea/Si substrate.d) 3D AFM image of N-graphene hydrangea/Si after five cleanings.e) Water contact angle change statistics of N-graphene hydrangea/Si after ethanol washing.f ) SERS spectrum of DEHP detected by N-graphene hydrangea/Si after washing.