General Synthesis of Large‐Area Transition Metal Nitride Porous Arrays for Highly Sensitive Surface‐Enhanced Raman Scattering Substrates with Ultrahigh Durability

A general polymerization‐induced strategy is developed to synthesize large‐area (100 cm2) transition metal nitride (TMN: TiN, MoN, WN, VN) porous arrays with high specific surface area (187.5–215.5 m2 g−1) for the first time, which also can be used to prepare TMN porous arrays uniformly doped with multiple elements. The formation of homogeneous organic/inorganic hybrid polymer precursors is a key factor in forming such TMN porous arrays. The TiN porous arrays exhibit an intense blue‐light localized surface plasmon resonance effect centered at 508 nm. The TiN porous arrays also show a remarkable surface‐enhanced Raman scattering (SERS) effect with a Raman enhanced factor of 5.7 × 107 and the lowest detection limit of 1.0 × 10−12 m. They show ultrahigh corrosion resistance and oxidation resistance, which are not available on traditional noble metal and semiconductor SERS substrates. These results suggest the possibility of the development of effective SERS substrates by using cheap TiN to replace the expensive commercial Au substrates.


Introduction
[3][4][5][6][7] However, TMNs, especially TMN porous arrays with high specific surface area and porosity, are very rare due to the thermodynamic limitations of their synthesis.20] Surface-enhanced Raman scattering (SERS) spectroscopy is becoming an extremely important analytical technique due to its high sensitivity and fingerprint resolution, and is widely used in fields such as risk substance identification, biological imaging, and chemical mechanism exploration. [21,22][25][26] Despite their many advantages, they also have shortcomings such as resource scarcity, unstable chemical properties, and high prices, which are determined by the nature of materials.Developing new, efficient, and stable substrate materials is an urgent issue in the application of SERS technology.Recently, some research groups have reported SERS substrates based on TiN nanostructures, such as disordered nanorod films, [27] porous nanotube arrays, [28] and nanoparticle films, [29] which provide reference for studying the SERS properties of TMN.However, despite the SERS effect exhibited by these TiN substrates, their Raman EF is low (10 3 -10 4 ) and cannot provide a highly sensitive detection limit, especially the signal uniformity of these substrates still falls far short of the standard for practical applications.More recently, we have synthesized large-area MoN flexible substrates and three-dimensional (3D) porous VN substrates, and found that they have an unexpected SERS effect. [30,31]However, their synthesis is constrained by environmentally unfriendly solvents and poorly preserved nitrogen sources, and their sensitivity as SERS substrates needs to be further improved.Therefore, DOI: 10.1002/sstr.202300127A general polymerization-induced strategy is developed to synthesize large-area (100 cm 2 ) transition metal nitride (TMN: TiN, MoN, WN, VN) porous arrays with high specific surface area (187.5-215.5 m 2 g À1 ) for the first time, which also can be used to prepare TMN porous arrays uniformly doped with multiple elements.The formation of homogeneous organic/inorganic hybrid polymer precursors is a key factor in forming such TMN porous arrays.The TiN porous arrays exhibit an intense blue-light localized surface plasmon resonance effect centered at 508 nm.The TiN porous arrays also show a remarkable surface-enhanced Raman scattering (SERS) effect with a Raman enhanced factor of 5.7 Â 10 7 and the lowest detection limit of 1.0 Â 10 À12 M. They show ultrahigh corrosion resistance and oxidation resistance, which are not available on traditional noble metal and semiconductor SERS substrates.These results suggest the possibility of the development of effective SERS substrates by using cheap TiN to replace the expensive commercial Au substrates.fabricating a large area of TMN-based SERS substrates by a relatively simple method and deeply revealing its Raman enhancement mechanism remains a major challenge.
Here, we report a general and facile polymerization-induced method to synthesize large-area (100 cm 2 ) TMN (TiN, MoN, WN, VN) porous array films with high specific surface area.The method is simple and general, and can not only be used to prepare pristine TMN porous array films, but also can be used to prepare TMN porous array films uniformly doped with multiple elements.This method can also be easily scaled up for large-scale preparation of TMN porous nanostructures at 300 g-level.The formation of homogeneous organic/inorganic hybrid polymer precursors is crucial in the formation of such TMN porous array films.The TiN porous array films exhibit strong and narrow blue-light LSPR effects centered at 508 nm.Furthermore, the TiN porous array films show a remarkable SERS effect with a Raman enhancement factor (EF) of 5.7 Â 10 7 and the lowest detection limit of 1.0 Â 10 À12 M. A synergistic Raman enhancement mechanism of LSPR and interface charge transfer (CT) is found in the TiN porous arrays.In addition, the TiN porous array films show ultrahigh corrosion resistance and oxidation resistance, which is lacking on traditional noble metal and semiconductor SERS substrates.

Results and Discussion
As shown in Figure 1, a general polymerization-inducing strategy is designed to synthesize large-area TMN porous array films.Taking the synthesis of TiN porous array films as an example, first, tetrabutyl titanate (TBT), furfuryl alcohol (FA), and a small amount of water are mixed together to form a dark-red transparent liquid.The dark-red solution gradually turned into a viscous liquid.The increase in viscosity can be attributed to the polymerization of FA molecules. [32]In the current case, Ti 4þ ions released from TBT act as a weak Lewis acid to mildly catalyze the polymerization of furfuryl alcohol monomers to form furfuryl alcohol resin (FAR) polymer.At the same time, Ti 4þ ions are closely combined with FAR through the coordination of oxygen-containing groups (such as hydroxyl and carbonyl) to form a uniform organic/inorganic hybrid polymer at the molecular level.This dark-red viscous liquid can be easily spin-coated onto glass or quartz substrate and form a uniform translucent film.As the polymerization degree increases, the dark-red translucent film gradually transforms into a black Ti/FAR film.Subsequently, the carbon component in the Ti/FAR film is completely removed by oxidation to obtain TiO 2 porous array films.Finally, TiN array films are obtained by in situ nitriding the TiO 2 porous array films.For the synthesis of other TMN (MoN, VN, NbN) porous array films, it only needs to replace the titanium source with the corresponding metal precursor (see the Experimental Section for detailed synthesis procedures).
We first characterized the as-synthesized TiO 2 array films (Figure S1a, Supporting Information).The X-Ray diffraction (XRD) pattern reveals that the crystalline phases of the white sample covered on the quartz plate can be indexed as anatase TiO 2 containing traces of rutile phase (Figure S1b, Supporting Information), [33] which is also confirmed by Raman spectrum (Figure S1c, Supporting Information). [34]Scanning electron microscope (SEM) images show that these TiO 2 grow vertically on the substrate surface with a morphology of microvillus.The microvilli are about 30 μm in length and 1.5-3 μm in diameter (Figure S1d-e, Supporting Information).Transmission electron microscope (TEM) images reveal that these TiO 2 microvilli are mesoporous, containing a large number of nanoparticles about 10 nm in size (Figure S1f, Supporting Information).This porous nature was also clearly demonstrated by a high-angle annular dark-field (HAADF) image (Figure S1g, Supporting Information).The selected area electron diffraction (SAED) patterns obtained at three different sites indicated highly consistent annular diffraction patterns, indicating that these particle scales are crystalline and highly uniform (Figure S1h, Supporting Information).High-resolution TEM (HRTEM) images further demonstrated that these TiO 2 microvilli are highly crystalline (Figure S1i, Supporting Information).
After nitriding, the white coating turns black (Figure 2a).SEM images show that the nitrided samples well retained the same microvillus-like morphology as TiO 2 (Figure 2b).The enlarged image shows that both the trunk (Figure 2c) and top (Figure 2d) of the arrays display well-defined mesoporous structures.TEM images further demonstrate that these nitrided samples are composed entirely of mesoporous structures (Figure 2e,f ).HRTEM image reveals that this porous structure is composed of highly crystalline nanoparticles (Figure 2g).These interplanar spacings are 0.21 and 0.24 nm, respectively, which match well with the ( 200) and (111) planes of cubic TiN. [35]The energy dispersive spectrum (EDS) mapping images show that Ti and N elements are uniformly distributed throughout the micropillar with a Ti/N ratio of 1.08 (Figure 2h), which is very close to that of TiN.XRD pattern (Figure 2i) revealed that these nitrided samples can be characterized as face-centered cubic TiN (PCPDS: 65-0715), which was also confirmed by their X-Ray photoelectron spectroscopy (XPS) spectra (Figure S2, Supporting Information) and Raman spectra (Figure S3, Supporting Information).In addition, nitrogen adsorption and desorption isotherms indicate the specific surface area of the TiN microvillus is up to 205.1 m 2 g À1 (Figure S4, Supporting Information).The above results demonstrate that large-area TiN porous arrays with high specific surface area have been successfully fabricated by this polymerization-inducing strategy.
The above experimental results show that the formation of large-scale TiO 2 porous array films is crucial for the later formed TiN porous array films.Formation mechanism investigation revealed that the use of TBT as a weak H þ ion donor to mildly catalyze the polymerization of FA monomers plays a key role in this large-scale TiO 2 porous array films.Traditionally, the rapid polymerization of FA monomer is usually initiated by strong acids such as H 2 SO 4 and HCl, [36] but such rapid polymerization is unfavorable to obtain the desired array structure.Comparative experiments show that if a small amount of HCl (2% by mass) is added to the reaction solution to speed up the polymerization reaction, the nitrided product is not TiN porous array films, but instead is a bulk porous structure composed of TiN nanoparticles (Figure S5, Supporting Information).
We scaled up the reaction to gain a deeper insight of the reasons for this difference.Expanding the volume of the reaction solution to 50 mL, after semipolymerization, a translucent jelly-like object with elasticity was obtained (Figure 3a).This jelly-like substance exhibits a pronounced Tyndall phenomenon, suggesting that it is uniformly composed of colloidal-scale microstructures.With the increase of the polymerization degree, the polymer gradually deepened to an opaque brown-black and hardened (Figure 3b, referred to as FAR-1).After the FAR-1 was oxidized in air, a large amount of TiO 2 powders were obtained (Figure 3c), whose crystal phase, morphology, and microstructure (Figure S6, Supporting Information) were highly consistent with those of TiO 2 porous array films grown on glass substrate.Keep the same volume, comparative experiments showed that the reaction solution with an additional addition of HCl (2% wt) rapidly polymerized into a non-transparent brown solid in only 30 min (Figure 3d), and transformed into a black hard solid within 3 h (Figure 3e, referred to as FAR-2).The TiO 2 formed by the oxidation of the FAR-2 is a massive mesoporous structure formed by random aggregation of nanoparticles (Figure 3f and S7, Supporting Information).
Based on the above results, there are reasons to believe that the high transparency and Tyndall effect are the external manifestations of the long-range order of the FAR-1 precursor.The slow release of H þ ions from TBT favors the formation of the ordered polymeric structure kinetically.Considering that the high-resolution mass spectrometry information of FAR-1 and FAR-2 is highly consistent (Figure 3g), it is reasonable to believe that the formation of this ordered polymeric structure is the key to the formation of TiO 2 porous microvillus arrays.Moreover, the thermogravimetric analysis (TGA) characterization shows that the phase transition from FAR-1 to TiO 2 requires higher temperatures compared to FAR-2 (Figure 3h), which suggests a kinetic insight of structural differences between FAR-1 and FAR-2.Although the precise formation mechanism is still under exploration, this novel polymerization-inducing strategy undoubtedly provides a simple and efficient method for the synthesis of largearea TiN porous arrays.Moreover, it is worth mentioning that this method can also be further scaled up for the preparation of TiO 2 and TiN porous nanostructures at the 300 g-scale (Figure 3i and S8, Supporting Information).This facile method is not only suitable for synthesizing large-scale TiN porous array films but also for synthesizing other important TMN porous array films, such as MoN, WN, and VN.The morphology and microstructure of the as-synthesized MoN, WN, and VN porous array films are investigated by SEM, HAADF, and HRTEM.SEM images show that all three samples exhibit a uniform pillar-array structure (Figure 4a,d,g).HAADF images reveal that all the samples are in the form of porous structures, and the size of grain of them is about 5-10 nm (Figure 4b,e,h).The HRTEM images reveal the lattice fringe spacing of 0.24, 0.21, and 0.24 nm, corresponding to hexagonal MoN (200), cubic WN (002) and cubic VN (111) crystal faces respectively (Figure 4c,f,i).Moreover, the successful synthesis of different types of TMNs is also confirmed by their XRD patterns, as shown in Figure S9, Supporting Information.All diffraction peaks of the three samples can be characterized as MoN (JCPDS No. 01-089-5024), [37] WN (JCPDS No. 75-1012), [38] and VN (JCPDS No. 35-0768). [6]EDS (Figure S10, Supporting Information) and XPS (Figure S11, Supporting Information) results also proved that MoN, WN, and VN were synthesized.Their specific surface areas are 187.5, 215.5, and 192.3 m 2 g À1 , respectively (Figure 4j-l).These comprehensive characterization results clearly demonstrate that this polymerization-inducing method is an effective and general way to synthesize TMN porous arrays.
Element doping can profoundly change the electronic structure of materials.The development of effective element doping methods is crucial for the development of high-performance materials.However, the preparation of doped samples at high temperature can easily lead to heteroatoms detached from the lattice and heterogenized.Due to the sintering phenomenon caused by high pressure and high temperature, the uniform doping of TMNs has always been an urgent problem to be solved.Fortunately, this general polymerization-inducted method can be used to prepare large-area TMN array films with homogeneous doping of multicomponents.Figure S5a,b, Supporting Information, presents the SEM and TEM images of the Mo-doped (1.0% wt) TiN porous array films, showing a homogeneous porous structure.In particular, the EDS mapping demonstrates that the Mo element is uniformly distributed throughout the TiN porous array films (Figure 5c).The effect of this method is also satisfactory for binary component doping.Figure 5d-e shows the TiN porous arrays codoped with Mo (1.0% by mass) and W (0.5% by mass).EDS mapping image demonstrates that the Mo and W components are uniformly distributed in the TiN porous array films (Figure 5f ).Moreover, the TiN array films uniformly doped with ternary components were successfully prepared by this method.Similar to the single-and two-component doping, the three-component TiN doped with Mo (1.0% by mass), W (0.5% by mass), and V (0.5% by mass) still retains the well-defined mesoporous structure (Figure 5g,h), and the distribution of the three components in the matrix is very homogeneous (Figure 5i).Due to the low concentration of doping components, the XRD patterns show that all three samples can be identified as pure cubic TiN (Figure S12, Supporting Information) with clear lattice fringes (Figure 5j-l), indicating that heteroatoms have entered the TiN matrix.These comprehensive characterization results convincingly demonstrate the significant advantages of the current polymerization-inducing method in preparing uniformly doped TMN porous nanostructures.We believe that the high viscosity and abundant coordination groups (Figure S13, Supporting Information) of FAR during the formation process effectively avoid the aggregation of dopant components, which is crucial in the formation of uniformly doped TiN.
The SERS effect of the TiN porous array films was investigated.As shown in Figure 6a, these TiN porous arrays with high specific surface area are ideal for SERS substrates, structurally, it is easy to form strong electromagnetic "hot spots" in the close-arranged mesopores, and the Raman scattering signals of analyte molecules adsorbed at these positions will be greatly enhanced under the stimulation of the excitation light.In tests, bisphenol A (BPA), a cancer-prone persistent environmental pollutant, was selected as a probe molecule to evaluate the SERS effect of the TiN porous arrays, and a 532 nm laser is used as the excitation light with a power of 0.7 mW.As shown in Figure 6b, through the detection of a series of BPA standard samples with different concentrations (10 À7 , 10 À9 , and 10 À11 M), these TiN porous arrays show highly sensitive Raman scattering effect.Furthermore, even if the concentration of analyte is diluted to 1.0 Â 10 À12 M, distinguishable Raman signals (S/N ≥ 4) can still be detected (Figure S14, Supporting Information), indicating the superior SERS properties of the TiN porous array films.It should be noted that this detection limit is better than most semiconductor SERS substrates, and even better than some noble-metal substrates. [39,40]By comparing the SERS signal intensity at 839.9 cm À1 of BPA sample (1.0 Â 10 À12 M) with the normal Raman signal intensity of 0.1 M BPA, the average value of the Raman EF of these TiN porous arrays is calculated to be as high as 5.7 Â 10 7 (Figure 6c).To the best of our knowledge, such high Raman EF may be one of the best non-noble-metal SERS records (Table S1, Supporting Information).
It is found that the strong Raman enhancement properties of the TiN porous array films are jointly dominated by the LSPR effect and the interface CT effect.The density functional theory (DFT) calculation reveals that face-centered cubic TiN is highly metallic.The region near its Fermi-level is composed of Ti 3d orbitals (Figure 6d), showing high-density free electrons.Moreover, the free electron gas (FEG) distribution simulated by electron local function (ELF) shows that the FEG density is high (Figure 6e).At 300 K, I-V profile indicates that the TiN exhibits high conductivity, and the linear temperature dependence of conductivity is highly consistent with typical metals (Figure 6f ). [4]More importantly, these TiN porous arrays exhibit unique blue-light LSPR effect centered at 508 nm (Figure 6g).It should be pointed out that for most of the reported TiN materials, their LSPR behaviors generally appear in the near-infrared (NIR) region or NIR/Vis junction area. [41,42]We believe that there are two main reasons for these different SPR effects.On the one hand, the reduction of the grain size can effectively reduce the recombination probability of electrons and holes, thereby increasing the electron density on the grain surface and causing the blue-shift of the LSPR effect. [43,44]On the other hand, the significantly blue-shifted LSPR peak may be also attributed to their unique close-packed mesoporous structure that strongly promotes the long-range dipole coupling. [45,46]This explanation is confirmed by comparative experimental results.After the TiN porous arrays were annealed at 1200 °C (Ar) for 2 h, their grain size would significantly from 5 to 10 nm increase to 30-50 nm, correspondingly, their LSPR effect was obviously red-shifted and weakened (Figure S15, Supporting Information).This strong blue-light LSPR effect provides an ideal physical enhancement basis for TiN porous arrays.
The distribution and intensity of electromagnetic hot spots on the surface of the TiN porous arrays by finite-difference time-domain (FDTD) near-field simulations were carried out.We stimulated the electromagnetic field distribution with the size value of TiN particles from around 5-10 nm and the gap value of particles from 1 to 2 nm, respectively.It can be seen from Figure 6h that the intensities of the hot spots at the gaps of the TiN nanoparticles are considerably strong under an excitation of 532 nm monochromatic light.The maximum value of E/E 0 can reach 27.4, which corresponds to 5.6 Â 10 5 field enhancement (E 4 ).The FDTD results demonstrate that the densely-arranged metallic TiN nanoparticles obviously have the mechanism of electromagnetically induced Raman signal enhancement.
However, this near-field Raman EF (5.6 Â 10 5 ) is almost two orders of magnitude weaker than the actual measured EF (5.7 Â 10 7 ), suggesting that there is also a chemical enhancement mechanism at work.Recently metallic MoTe 2 and WTe 2 nanosheets with atomic layer thickness were reported to be an outstanding non-noble metal SERS substrate with an EF of 10 8 -10 9 , and the ultrahigh SERS activity was attributed to the high adsorption energy (E ad = 0.67 eV) and interface CT effect between MoTe 2 (WTe 2 ) and analyte molecules (R6G). [47]ased on these early findings, we can reasonably assume that the chemical enhancement also occurs between the BPA molecule and the metallic TiN substrate, since the highly crystalline TiN nanoparticles are only 5-10 nm in size.DFT calculations show that the E ad between BPA molecule and TiN (001 plane) is up to 0.465 eV, which indicates that there is a strong interfacial interaction between BPA and TiN nanoparticles (Figure 6i).Furthermore, by using Bader's quantum theory of atoms in molecules (QTAIM) charge analysis, there are 0.47 e per unit transferred from BPA molecule to TiN nanoparticles.[50] Experimentally, the UV-vis spectrum shows that when BPA is adsorbed on these TiN porous arrays, its absorption spectrum changes significantly, and new absorption bands appear (Figure S16, Supporting Information), which experimentally shows that BPA and TiN substrate significant interface CT behavior does occur between them.
Three persistent environmental pollutants with strong carcinogenicity, dichlorophenol (2,4-DCP), trichlorophenol (2,4,5-TCP), and tetrachlorophenol (2,3,4,6-TeCP), are used as probes to evaluate the universality of the TiN porous array substrate.In all tests, unless otherwise specified, the excitation light power is 0.7 mW, the spot diameter is 5 μm, the excitation wavelength is 532 nm, and the excitation light irradiated the substrate perpendicularly.The test results show that it exhibits strong SERS effect on all three molecules (Figure 7a-c).Through the detection of a series of analytes with different concentrations, it was found that these TiN substrates have the lowest detection limit of 1 Â 10 À12 M for the three molecules.The signal uniformity of the TiN substrate was also investigated.The signal intensities recorded at 5000 randomly selected measurement points were statistically used to calculate their relative standard deviation (RSD).As shown in Figure 7d-f, the statistic shows that for 10 À10 M 2,4-DCP, 2,4,5-TCP, and 2,3,4,6-TeCP, the calculated RSD values are 7.8%, 6.7%, and 8.2%, respectively, which is significantly better than the commercial Au substrates (%20%).Furthermore, the outstanding advantage over noble-metal and semiconductor SERS substrates is that the TiN microvillus exhibits ultrahigh oxidation resistance and corrosion resistance.Even after harsh chemical treatments such as strong acid (in 5 M HCl for 5 h), strong base (in 5 M NaOH for 5 h), or high temperature (in the air at 300 °C for 3 h), these WN/C nanobelts still show almost unchanged SERS activity (Figure 7h-j) and microstructures (Figure S17-19, Supporting Information), indicating their potential for application in extreme environments.

Conclusion
In summary, a simple and versatile polymerization-induced strategy is developed to fabricate large-area (100 cm 2 ) TMN porous array films with high specific surface area for the first time.Highly homogeneous multicomponent-doped TMN porous arrays can also be synthesized by this method.This method can also be easily scaled up for 300 g-scale preparation of TMN porous nanostructures.The TiN porous arrays exhibit strong blue-light LSPR effect and remarkable SERS effect.The current results provide a reference method for fabricating large-area TMN porous arrays, which provide avenues for discovering their new properties and applications.

Experimental Section
Synthesis of Large-Area TiO 2 Porous Arrays: In a typical synthesis, 5 mL of TBT and 0.5 mL of deionized water were added to 30 mL of FA, stirred for 10 min, and left to stand for 4 h at 50 °C in air.The resulting dark-yellow viscous liquid was coated on a 10 Â 10 cm quartz plate by spin-coating and aged to brown-black in air at 90 °C.Finally, the brown-black quartz plate was placed in a high-temperature atmosphere furnace (air), heated to 500 °C at a heating rate of 1 °C min À1 , and held at this temperature for 1 h.After the reaction, it was naturally cooled to room temperature to obtain a quartz plate covered with TiO 2 porous arrays.
Synthesis of Large-Area TiN Porous Arrays: The previously synthesized quartz plates covered with TiO 2 microvillus arrays were placed in a high-temperature atmosphere furnace.Before heating, the furnace chamber was purged of air with N 2 .The atmosphere furnace was heated to 600 °C at a heating rate of 1 °C min À1 and held for 2 h.During the heating period, N 2 was used as the carrier gas (2 mL s À1 ) and NH 3 was used as the nitrogen source (3 mL s À1 ), respectively.
SERS Test: To study the SERS properties of these as-synthesized TiN porous arrays, a confocal micro Raman spectrometer (Renishaw-inVia Qontor) was used as the measuring instrument.In all SERS tests, unless specifically stated, the excitation wavelength is 532 nm (the selection of excitation wavelength should follow the principle of electromagnetic resonance: the excitation wavelength and substrate absorption wavelength should be as close as possible), laser power is 0.7 mW and the specification of the objective is Â 50 L.A series of standard ethanol solution of 2,4-DCP, 2,4,5-TCP, and 2,3,4,6-TeCP with concentrations of 10 À7 -10 À12 M are used as the probe molecules.To improve the signal reproducibility and uniformity, the quartz plates covered with TiN porous arrays were immersed into a probe solution (30 mL) to be measured for 3 min, then taken out and dried in air for 5 min.In all SERS tests, the laser beam is perpendicular to the top of the sample to be tested with a resultant beam spot diameter of 5 μm.The fluorescent background of the probe molecule is deducted by the software that comes with the instrument.
Material Characterization: These TiN porous arrays are measured by a variety of characterization techniques.XRD patterns of the products were obtained on a Bruker D8 focus X-Ray diffractometer by using CuKa radiation (l = 1.54178Å).SEM images and EDS were obtained on a Hitachi S-4800.TEM, HRTEM, EDS, and SAED characterizations are performed with a JEOL F200 operated at 200 kV.UV-Vis absorption spectra are recorded with a Shimadzu UV3600-Plus.The specific surface area and pore size were measured in a Micro Tristar II 3020.Raman spectra were recorded from Renishaw-inVia Qontor.

Figure 1 .
Figure 1.Schematic diagram of the synthesis of the large-area TiN porous array film.a) Schematic diagram of each stage.b) Physical images of each stage.

Figure 2 .
Figure 2. Morphology and structural characterization of the large-area TiN microvillus arrays.a) Optical photos of the TiN film.b-d) SEM images.e-f ) TEM images.g) HRTEM image.h) EDS spectrum and mapping image.i) XRD pattern.

Figure 3 .
Figure 3. Factors influencing TiN microvillus arrays formation.a-c) The process of TBT catalyzes the polymerization of FA into FAR.d-f ) The process of strong acid-accelerated polymerization of FA to FAR; g) High-resolution mass spectra of the FAR-1 and FAR-2.h) TG curves of the FAR-1 and FAR-2.i) TiN microvillus obtained by large-scale synthesis.

Figure 4 .
Figure 4. Morphology and structure characterization of the TMN porous array films.a-c) MoN.d-f ) WN. g-i) VN. j-l) N 2 adsorption and desorption of the j) MoN porous array films, k) WN porous array films, and l) VN porous array films.

Figure 5 .
Figure 5. Morphology, structure, and element distribution characterization of the doped TiN microvillus arrays.a-c) Mo-TiN.d-f ) Mo (W)-TiN.g-i) Mo (W,V)-TiN.j-l) HRTEM images of the doped TiN arrays with different dopants.

Figure 6 .
Figure 6.SERS effect and enhancement mechanism of TiN porous array films.a) Schematic diagram of the TiN porous array SERS substrate.b) Response of the substrate to different concentrations of BPA.Excitation wavelength: 532 nm.Integration time: 2 s for 10 À7 M, 3 s for 10 À9 M, 5 s for 10 À11 M. c) EFs of the TiN porous array substrate at different concentrations.d) The band structure of cubic phase TiN.e) The ELF of TiN.f ) I-V curves of the TiN porous array films at 300 K. g) UV-vis absorption of the TiN porous array films.h) Hot spots simulated by FDTD.i) Side views of the electron density differences BPA adsorbed on TiN porous array films.