Preparation of Self-Assembled Composite Films Constructed by Chemically-Modified MXene and Dyes with Surface-Enhanced Raman Scattering Characterization

The effective functionalization and self-assembly of MXene are of crucial importance for a broad range of nanomaterial applications. In this work, we investigated the aggregates of sulfanilic acid-modified MXene (abbreviated as MXene-SO3H) with three model dyes at the air–water interface and demonstrated the morphological and aggregation changes of composite films, using Langmuir-Blodgett (LB) technology, as well as excellent uniformity and reproducibility by using surface-enhanced Raman scattering (SERS) spectra. This research has found that cationic dye molecules were adsorbed onto negatively charged MXene-SO3H particles mainly through electrostatic interaction and the particles induced dyes to form highly ordered nanostructures including H- and/or J-aggregates corresponding to monomers in bulk solution. The surface pressure-area isotherms from different dye sub phases confirmed that the stable composite films have been successfully formed. And the spectral results reveal that different dyes have different types of aggregations. In addition, the SERS spectra indicated that the optimal layers of MXene-SO3H/methylene blue (MB) films was 50 layers using rhodamine 6G (R6G) as probe molecule. And the formed 50 layers of MXene-SO3H/MB films (MXene-SO3H/MB-50) as SERS substrate were proved to possess excellent uniformity and repeatability.


Introduction
The aggregation of dye molecules plays a crucial role in both human life and industrial engineering as well as material sciences. The performance of material relies not only on the molecular structure that constitutes it but also on its state of aggregation. For examples, Zhang et al. designed a reversible dye film switch [1]. The formed dye aggregates in the films could be reversibly switched by treating the cyanine dyes and gemini amphiphiles with rigid spacers and complex films with

Materials
MXene (Ti 3 C 2 Tx) was prepared by adding Ti 3 AlC 2 powders into aqueous solution of 12 M LiF and 9 M HCl for 24 h to move an Al component in a polytetrafluoroethylene beaker. Then the suspension was washed using deionized water and centrifuged to separate the powders from the supernatant. After alkaline treatment with dilute NaOH solution for 2 h, the sample was washed by the deionized water and then placed in a vacuum oven at 60 • C to dry for 24 h. Finally, the products were treated by vacuum calcination at 600 • C [35,36]. Sulfanilic acid diazonium salts were synthetized according to literatures [20,37,38]. Methylene blue (MB) was purchased from Aladdin Reagent (Shanghai, China) and Alfa Aesar Chemicals (Tianjin, China). Rhodamine B (RhB), SafranineT (ST) and sodium hydroxide (NaOH) were obtained from Tianjin KaiTong Chemical Reagent (Tianjin, China). Methanol was analytical reagent grade. All chemicals were used without further purification and deionized (DI) water was used in all experiments. R6G was purchased from J&K scientific LTD.

Surface Modification of MXene
The used MXene-SO3H material was prepared as described in a previous report [20]. The specific experimental method is as follows: 0.05 g MXene powders were immersed in 100 mL deionized water under ultrasonication; 5 g sodium hydroxide was stir-mixed with MXene suspension for 2 h; then the suspension was centrifuged following a big amount of DI water until pH = 7. Subsequently, the obtained Na + intercalated MXene was chemical modification. The intercalated MXene was immersed in 10 mL DI water and next stir-mixed with above sulfanilic acid diazonium salts for around 6 h under ice-water mixture bath and then centrifuged for 15 min at 8000 rpm. The collected sediment was rinsed with DI water and subsequently centrifuged for 15 min at 5000 rpm to remove unexfoliated particles. The obtained nanoscale surface-modified MXene (abbreviated as MXene-SO3H) dispersion solution was filtered using 0.45 um membrane and finally mild ultrasonication before lyophilized to obtain the powder. Figure 1a demonstrated the measured lateral size of the particles was around 100 nm. The inserted photo represented a colloidal solution of MXene-SO3H particles that exhibited the Tyndall scattering effect when the beam passed through. Figure 1b demonstrated clearly that the representative MXene particles were thin and evenly distributed. The measured high-resolution TEM image of MXene-SO3H showed obvious crystal layer spacing along the c-axis about 1.51 nm. In addition, the zeta potential of MXene-SO3H was determined as −31.2 mV at a neutral pH of 7.0, which seemed similar to previous reported value (−30.6 mV) [20], indicating the good dispersibility of MXene-SO3H particles in water. Tyndall scattering effect and the high-resolution TEM image of MXene-SO3H.

Preparation of Composite Langmuir Films
For the interfacial preparation and the deposited multilayer films, a commercially available KSV-NIMA LB trough (KN 2002, Biolin Scientific, Stockholm, Sweden) was carefully cleaned with ethanol as well as water and then filled with pure water, fresh stock MB, RhB or ST solutions with concentrations of 10 −3 mol/L to serve as sub phases, respectively. It was worth noting that the word "layer" films represented the performed number of transfer layers and were not related to MXene "layer." The as-prepared MXene-SO3H powder was dispersed in methanol/water (2/1 of v/v) as a spreading solvent to form stable dispersion and then 50 μL MXene-SO3H aqueous dispersion (0.7 mg/mL) was dropwise spread onto the sub phase surface by using a glass syringe at room temperature. Surface pressure was monitored by a tensiometer attached to a Wilhelmy plate. The spread film was left for about 20 min and the solvent (methanol) was evaporated to form a stable film before the barriers took place isothermal compression at a speed of 8 mm/min. When the monolayer film got compressed, a deep color layer at air-water interface was observed. In this process, surface pressure-area isotherms could be obtained. The prepared MXene-SO3H/dye composite films were also deposited onto different solid substrates (fresh cleaved mica, quartz, glass and CaF2 plates for AFM, UV-vis, SEM and FT-IR spectral measurements, respectively) at pressure of 15 mN/m by vertically dipping the substrate into the trough and slowly pulling it up (2 mm/min) for next morphological and spectral characterization.

SERS Measurements
For the SERS measurement, all of the Raman spectra were measured by the Renishaw inVia Raman microscope with the 532 nm laser as excitation source. For the laser, the diameter of the light spot was ~1 μm and the incident power was 0.05 mW. In order to compare the SERS performance, four different layer substrates have been prepared to select the best substrate by R6G solution (10 −3 M, 10 μL). The SERS performance such as uniformity and reproducibility was measured by the best substrate. In addition, the spectra were recorded by the 2 accumulations, the 10 s exposure time and the ×50 objective.

Characterization
The microstructure was characterized via transmission electron microscopy (TEM, HT7700, High-Technologies Corp., Ibaraki, Japan). High-resolution transmission electron microscopy (HRTEM, Tecnai-G 2 F30 S-TWIN, Philips, Netherlands) were used to observe the morphologies and microstructures of the samples. Atomic force microscopy (AFM) measurements were carried out with a Nanoscope model Multimode 8 Scanning Probe Microscope (Veeco Instrument, Santa Barbara, CA,

Preparation of Composite Langmuir Films
For the interfacial preparation and the deposited multilayer films, a commercially available KSV-NIMA LB trough (KN 2002, Biolin Scientific, Stockholm, Sweden) was carefully cleaned with ethanol as well as water and then filled with pure water, fresh stock MB, RhB or ST solutions with concentrations of 10 −3 mol/L to serve as sub phases, respectively. It was worth noting that the word "layer" films represented the performed number of transfer layers and were not related to MXene "layer." The as-prepared MXene-SO3H powder was dispersed in methanol/water (2/1 of v/v) as a spreading solvent to form stable dispersion and then 50 µL MXene-SO3H aqueous dispersion (0.7 mg/mL) was dropwise spread onto the sub phase surface by using a glass syringe at room temperature. Surface pressure was monitored by a tensiometer attached to a Wilhelmy plate. The spread film was left for about 20 min and the solvent (methanol) was evaporated to form a stable film before the barriers took place isothermal compression at a speed of 8 mm/min. When the monolayer film got compressed, a deep color layer at air-water interface was observed. In this process, surface pressure-area isotherms could be obtained. The prepared MXene-SO3H/dye composite films were also deposited onto different solid substrates (fresh cleaved mica, quartz, glass and CaF 2 plates for AFM, UV-vis, SEM and FT-IR spectral measurements, respectively) at pressure of 15 mN/m by vertically dipping the substrate into the trough and slowly pulling it up (2 mm/min) for next morphological and spectral characterization.

SERS Measurements
For the SERS measurement, all of the Raman spectra were measured by the Renishaw inVia Raman microscope with the 532 nm laser as excitation source. For the laser, the diameter of the light spot was~1 µm and the incident power was 0.05 mW. In order to compare the SERS performance, four different layer substrates have been prepared to select the best substrate by R6G solution (10 −3 M, 10 µL). The SERS performance such as uniformity and reproducibility was measured by the best substrate. In addition, the spectra were recorded by the 2 accumulations, the 10 s exposure time and the ×50 objective.

Characterization
The microstructure was characterized via transmission electron microscopy (TEM, HT7700, High-Technologies Corp., Ibaraki, Japan). High-resolution transmission electron microscopy (HRTEM, Tecnai-G 2 F30 S-TWIN, Philips, Netherlands) were used to observe the morphologies Nanomaterials 2019, 9, 284 5 of 19 and microstructures of the samples. Atomic force microscopy (AFM) measurements were carried out with a Nanoscope model Multimode 8 Scanning Probe Microscope (Veeco Instrument, Santa Barbara, CA, USA) to analyze the morphologies of the sample surface. The root-mean-square (rms) roughness of the obtained composite films was examined from the AFM images with a size of 10 × 10 µm 2 . FT-IR spectra was measured via a Fourier infrared spectroscopy (Thermo Nicolet Corporation, Madison, WI, USA) using the KBr tablet method. UV-vis spectra were obtained with a Shimadzu UV-2550 system (Shimadzu Corporation, Kyoto, Japan). Raman spectra for the experiment were measured by confocal micro-Raman spectrometer (inVia). The size distribution and zeta potential of present material was analyzed with the Nanozetasizer machine (ZEN 3690, Malvern Instruments, Malvern, UK). We obtained X-ray photoelectron spectroscopy (XPS) data by monitoring a Thermo Scientific ESCALab 250Xi (Netzsch Instruments Manufacturing Co., Ltd., Seligenstadt, Germany) equipped with 200 W of monochromatic AlKα radiation.

Results and Discussion
First, when the used spreading solvent volume was 50 µL, the used different spreading mixed solvents for MXene-SO3H Langmuir films spread on RhB sub phase surface is shown in Figure 2a.
The obtained result demonstrated that the optimized spread condition was methanol/water (v/v) of 2:1 and applied to all the following experiments. A representative surface pressure-area (π-A) isotherms of amphiphilic MXene-SO3H spread on pure water surface and the aqueous dye sub phases containing 10 −3 mol/L MB, RhB or ST at room temperature, respectively, as displayed in Figure 2b. The highest surface pressure of MXene-SO3H spread on pure water surface was about 2 mN/m, indicating that the as-prepared MXene-SO3H could not form stable film on the surface of pure water by self-assembly which probably due to the dissolution in water. In addition, obvious differences were observed for the isotherms with dye molecules dissolved in sub phase solutions. Obvious increment of surface pressure suggested the formation of MXene-SO3H/dye composite films due to reasonable intermolecular weak forces, such as π-π stacking, hydrogen bond and electrostatic interaction between dyes and MXene-SO3H particles, which reduced the electrostatic repulsion among MXene-SO3H particles and resulted in the formation of dense and uniform composite films [10].
Nanomaterials 2017, 7, x FOR PEER REVIEW 5 of 18 USA) to analyze the morphologies of the sample surface. The root-mean-square (rms) roughness of the obtained composite films was examined from the AFM images with a size of 10 × 10 μm 2 . FT-IR spectra was measured via a Fourier infrared spectroscopy (Thermo Nicolet Corporation, Madison, WI, USA) using the KBr tablet method. UV-vis spectra were obtained with a Shimadzu UV-2550 system (Shimadzu Corporation, Kyoto, Japan). Raman spectra for the experiment were measured by confocal micro-Raman spectrometer (inVia). The size distribution and zeta potential of present material was analyzed with the Nanozetasizer machine (ZEN 3690, Malvern Instruments, Malvern, UK). We obtained X-ray photoelectron spectroscopy (XPS) data by monitoring a Thermo Scientific ESCALab 250Xi (Netzsch Instruments Manufacturing Co., Ltd., Seligenstadt, Germany) equipped with 200 W of monochromatic AlKα radiation.

Results and Discussion
First, when the used spreading solvent volume was 50 μL, the used different spreading mixed solvents for MXene-SO3H Langmuir films spread on RhB sub phase surface is shown in Figure 2a.
The obtained result demonstrated that the optimized spread condition was methanol/water (v/v) of 2:1 and applied to all the following experiments. A representative surface pressure-area (π-A) isotherms of amphiphilic MXene-SO3H spread on pure water surface and the aqueous dye sub phases containing 10 −3 mol/L MB, RhB or ST at room temperature, respectively, as displayed in Figure  2b. The highest surface pressure of MXene-SO3H spread on pure water surface was about 2 mN/m, indicating that the as-prepared MXene-SO3H could not form stable film on the surface of pure water by self-assembly which probably due to the dissolution in water. In addition, obvious differences were observed for the isotherms with dye molecules dissolved in sub phase solutions. Obvious increment of surface pressure suggested the formation of MXene-SO3H/dye composite films due to reasonable intermolecular weak forces, such as π-π stacking, hydrogen bond and electrostatic interaction between dyes and MXene-SO3H particles, which reduced the electrostatic repulsion among MXene-SO3H particles and resulted in the formation of dense and uniform composite films [10]. In order to assess the nanostructures of the prepared MXene-SO3H/dye films, TEM and AFM techniques were utilized, as shown in Figure 3. AFM images of morphology and root-mean-squared (rms) roughness were performed by means of transferring the films to mica sheet at surface pressure of 15 mN/m. For present obtained MXene-SO3H/dye composite films, flat film morphology with some aggregates could be observed in TEM images, as shown in Figure 3a-c. At the same time, it should be noted that present MXene-SO3H material synthesized via centrifuge at range of 5000-8000 rpm showed particles lateral size of around 100 nm. In addition, AFM images in Figure 3d-f showed many little aggregate domains in the prepared MXene-SO3H/dye composite films, suggesting the In order to assess the nanostructures of the prepared MXene-SO3H/dye films, TEM and AFM techniques were utilized, as shown in Figure 3. AFM images of morphology and root-mean-squared (rms) roughness were performed by means of transferring the films to mica sheet at surface pressure of 15 mN/m. For present obtained MXene-SO3H/dye composite films, flat film morphology with some aggregates could be observed in TEM images, as shown in Figure 3a-c. At the same time, it should be noted that present MXene-SO3H material synthesized via centrifuge at range of 5000-8000 rpm showed particles lateral size of around 100 nm. In addition, AFM images in Figure 3d-f showed many little aggregate domains in the prepared MXene-SO3H/dye composite films, suggesting the reasonable aggregations of dye molecules with MXene, which demonstrated similar interfacial self-assembly process of graphene oxide-dye composite films reported previously [10]. For example, Ding et al. achieved advanced water purification membranes using MXene with lateral size of 100-400 nm, which revealed excellent stability and high rejection rate (90%) for molecules with sizes larger than 2.5 nm [39]. Moreover, root-mean-squared (rms) roughness demonstrated the values of 1.93, 5.21 and 0.81 for MXene-SO3H/MB films, MXene-SO3H/RhB films and MXene-SO3H/ST films, respectively, as shown in Figure 3d'-f'. The higher roughness for the MXene-SO3H/RhB films could be explained by the denser adsorption of RhB molecules on MXene. Moreover, the difference of patterns of the obtained Langmuir films from different dye sub phase solutions were investigated, which could be mainly owing to the structural differences of dyes molecules. In addition, the elemental mapping scans and line scans of MXene-SO3H/RhB composite films in TEM measurement were demonstrated in Figures 4 and 5, respectively. The results of N/S elemental scans further confirmed the presence and the good distribution of RhB molecules on surface of MXene. It could be reasonably speculated that the used dye molecules successfully loaded on the surface of MXene-SO3H particles in composite films via intermolecular electrostatic forces and π-π stacking, which could be expected to exert good stability in the next application process. Figure 6a shows XPS profiles of the obtained MXene-SO3H/RhB composite films. In addition, as shown in Figure 6b, the peak at 167.8 eV was attributed to sulfonic acid groups in the S2p XPS spectrum, while the peak at 166 eV was assigned to the sulfoxides [20,40]. Moreover, the atomic percentages of MXene-SO3H/RhB composite films from XPS were shown in Table 1. It could be reasonably speculated that C and S elements came mainly from MXene-SO3H and the dyes molecules with the atomic contents of 71.30% and 2.24%, which also suggested that phenylsulfonic groups were successfully grafted onto the surface of MXene layers. Meanwhile, the presence of N element with the atomic contents of 5.53% mainly originated from MB molecules anchored on MXene-SO3H surface rather than contaminants from surrounding. These above obtained results clearly indicated successful preparation of composite films from dye molecules with MXene-SO3H.          Next, the composite multilayer MXene-SO3H/dye films were transferred to solid substrate and characterized by UV-vis spectra to investigate the aggregation of dye molecules in floating-layers, as shown in Figure 7. All the composite multilayer films were transferred by using horizontal lifting method at surface pressure of 15 mN/m. It could be clearly observed that the MXene-SO3H solution showed the obvious absorption peak around 227 nm. The MB solution showed a strong band at 664 nm together with a shoulder peak at 618 nm, which could be assigned to the monomer and Haggregate, respectively, as shown in Figure 7a. In addition, the composite films of MXene-SO3H/MB showed the absorption maximum around 605 nm with a relatively weak shoulder peak around 676 nm. The former band could be ascribed to the H-aggregation of the MB dye molecules, whereas the peak at 676 nm was related to the absorption of dimer or oligomer. This result indicated that the dye molecules adsorbed on the surface of Mxene-SO3H formed ordered aggregates. Similar spectral change could be observed in Figure 7b. RhB aqueous solution showed an intense narrow band at 553 nm, with a shoulder peak around 520 nm, suggesting the dye molecules mainly existing in the form of monomers in aqueous solution. However, a redshifted band at 566 nm could be attributed to the J-aggregation in obtained MXene-SO3H/RhB composite films. In addition, a shoulder peak was also observed at 539 nm, which was in accordance with a smaller H-aggregate of RhB molecules. In comparison, an absorption peak located at 518 nm of ST solution was observed in Figure 7c. While absorption peak in MXene-SO3H/ST multilayers took place blue shift, suggesting that ST dye molecules assembled onto the surface of functionalized MXene sheets in the form of H-aggregation. These results clearly indicated that the used model dyes could be orderly arranged onto the surface of MXene-SO3H to form different aggregate states mainly through electrostatic interactions and other cooperative forces.
To further characterize the obtained MXene materials and successful preparation of composite films, the FT-IR spectra were also measured and shown in Figure 8, which could be speculated to provide more information about the grafting of sulfonic acid groups and dye molecules successfully adsorbed on surface of MXene-SO3H. Compared with the pristine MXene, surface-modified MXene showed characteristic bands at 1189 and 1009 cm −1 , which could be assigned to the p-substituted phenyl groups characteristic vibrations [20,41,42]. The peaks located at 1124 cm −1 and 1038 cm −1 were contributed to S=O bond and S-phenyl band vibrations and compared to S=O bond at 1176 cm −1 from sulfanilic acid, it showed a slightly blue shift, indicating phenylsulfonic groups were successfully grafted on the MXene-SO3H surface. Other bands at 1635 cm −1 , 1591 cm −1 , 1598 cm −1 and 1608 cm −1 could be assigned to the skeletal vibrations of benzene ring of MXene-SO3H, RhB, MB and ST components. It could be easily observed that compared to MXene and MXene-SO3H materials, Next, the composite multilayer MXene-SO3H/dye films were transferred to solid substrate and characterized by UV-vis spectra to investigate the aggregation of dye molecules in floating-layers, as shown in Figure 7. All the composite multilayer films were transferred by using horizontal lifting method at surface pressure of 15 mN/m. It could be clearly observed that the MXene-SO3H solution showed the obvious absorption peak around 227 nm. The MB solution showed a strong band at 664 nm together with a shoulder peak at 618 nm, which could be assigned to the monomer and H-aggregate, respectively, as shown in Figure 7a. In addition, the composite films of MXene-SO3H/MB showed the absorption maximum around 605 nm with a relatively weak shoulder peak around 676 nm. The former band could be ascribed to the H-aggregation of the MB dye molecules, whereas the peak at 676 nm was related to the absorption of dimer or oligomer. This result indicated that the dye molecules adsorbed on the surface of Mxene-SO3H formed ordered aggregates. Similar spectral change could be observed in Figure 7b. RhB aqueous solution showed an intense narrow band at 553 nm, with a shoulder peak around 520 nm, suggesting the dye molecules mainly existing in the form of monomers in aqueous solution. However, a redshifted band at 566 nm could be attributed to the J-aggregation in obtained MXene-SO3H/RhB composite films. In addition, a shoulder peak was also observed at 539 nm, which was in accordance with a smaller H-aggregate of RhB molecules. In comparison, an absorption peak located at 518 nm of ST solution was observed in Figure 7c. While absorption peak in MXene-SO3H/ST multilayers took place blue shift, suggesting that ST dye molecules assembled onto the surface of functionalized MXene sheets in the form of H-aggregation. These results clearly indicated that the used model dyes could be orderly arranged onto the surface of MXene-SO3H to form different aggregate states mainly through electrostatic interactions and other cooperative forces.
To further characterize the obtained MXene materials and successful preparation of composite films, the FT-IR spectra were also measured and shown in Figure 8, which could be speculated to provide more information about the grafting of sulfonic acid groups and dye molecules successfully adsorbed on surface of MXene-SO3H. Compared with the pristine MXene, surface-modified MXene showed characteristic bands at 1189 and 1009 cm −1 , which could be assigned to the p-substituted phenyl groups characteristic vibrations [20,41,42]. The peaks located at 1124 cm −1 and 1038 cm −1 were contributed to S=O bond and S-phenyl band vibrations and compared to S=O bond at 1176 cm −1 from sulfanilic acid, it showed a slightly blue shift, indicating phenylsulfonic groups were successfully grafted on the MXene-SO3H surface. Other bands at 1635 cm −1 , 1591 cm −1 , 1598 cm −1 and 1608 cm −1 could be assigned to the skeletal vibrations of benzene ring of MXene-SO3H, RhB, MB and ST components. It could be easily observed that compared to MXene and MXene-SO3H materials, benzene ring skeleton stretches of composite films demonstrated a blue shift, suggesting more stronger electrostatic interaction between MXene-SO3H and dye molecules [43]. RhB showed a typical C=O at 1711 cm −1 [44]. In addition, compared to MXene-SO3H, new peaks of imine band C=N at 1648 cm −1 , 1650 cm −1 and 1645 cm −1 , C-N bond at 1181, 1176, 1190 cm −1 as well as methyl bond at 1339, 1335, 1331 cm −1 appeared on RhB, MB, ST curves, respectively [45][46][47]. And the peaks of C-O, C-H appeared at about 1180 cm −1 and 1390 cm −1 , respectively. ST and RhB molecules exhibited stretching vibration band belonging to N-H at 1532 cm −1 and C-O-C at 1076 cm −1 [48]. In addition, the new peaks at 3344 and 3206 cm −1 suggested different strong hydrogen bonding in the MXene-SO3H/ST composite films. These above results indicated that dye molecules existed on the surface of MXene-SO3H.
benzene ring skeleton stretches of composite films demonstrated a blue shift, suggesting more stronger electrostatic interaction between MXene-SO3H and dye molecules [43]. RhB showed a typical C=O at 1711 cm −1 [44]. In addition, compared to MXene-SO3H, new peaks of imine band C=N at 1648 cm −1 , 1650 cm −1 and 1645 cm −1 , C-N bond at 1181, 1176, 1190 cm −1 as well as methyl bond at 1339, 1335, 1331 cm −1 appeared on RhB, MB, ST curves, respectively [45][46][47]. And the peaks of C-O, C-H appeared at about 1180 cm −1 and 1390 cm −1 , respectively. ST and RhB molecules exhibited stretching vibration band belonging to N-H at 1532 cm −1 and C-O-C at 1076 cm −1 [48]. In addition, the new peaks at 3344 and 3206 cm −1 suggested different strong hydrogen bonding in the MXene-SO3H/ST composite films. These above results indicated that dye molecules existed on the surface of MXene-SO3H.  benzene ring skeleton stretches of composite films demonstrated a blue shift, suggesting more stronger electrostatic interaction between MXene-SO3H and dye molecules [43]. RhB showed a typical C=O at 1711 cm −1 [44]. In addition, compared to MXene-SO3H, new peaks of imine band C=N at 1648 cm −1 , 1650 cm −1 and 1645 cm −1 , C-N bond at 1181, 1176, 1190 cm −1 as well as methyl bond at 1339, 1335, 1331 cm −1 appeared on RhB, MB, ST curves, respectively [45][46][47]. And the peaks of C-O, C-H appeared at about 1180 cm −1 and 1390 cm −1 , respectively. ST and RhB molecules exhibited stretching vibration band belonging to N-H at 1532 cm −1 and C-O-C at 1076 cm −1 [48]. In addition, the new peaks at 3344 and 3206 cm −1 suggested different strong hydrogen bonding in the MXene-SO3H/ST composite films. These above results indicated that dye molecules existed on the surface of MXene-SO3H.   Figure 9 showed the obvious structural differences of three used dye molecules including MB, RhB and ST together with the top and side view of their space-filling models. The common feature of these three dyes was that they all contained large heterocyclic conjugated systems. The dimensions of the organic dyes are defined as x × y × z, wherein x, y and z represent the three edges of a cuboid that can just include the molecule. MB is one kind of planar molecules with x value of 16.3 Å, y value of 7.9 Å and z value of 4.0 Å. RhB is cationic with a triangle structure and contains a carboxylic group, wherein the values of x, y, z are 17.9 Å, 9.8 Å and 4.3 Å, respectively. In addition, ST demonstrates a T-type structure with x value of 16.1 Å, y value of 9.6 Å and z value of 4.3 Å [49][50][51][52][53][54][55]. Combined with the above results, schematic illustration of the preparation process of MXene-SO3H and the self-assembled modes in MXene-SO3H/dye composite Langmuir films were shown in Figure 10. At first, the targeted amphiphilic MXene-SO3H were prepared by intercalation of Na + ions and the modification of sulfanilic acid diazonium salts. Then the MXene-SO3H/dye composites Langmuir films and interfacial self-assembled behaviors at the air-water interface have been characterized and investigated via π-A isotherms, AFM and FT-IR techniques. Depending on different molecular structures in three used model dyes, H-and/or J-aggregates could be formed in different MXene-SO3H composite LB films, demonstrating reasonable and precise regulation of composite films via design of suitable components in organized nanostructures [56][57][58][59][60][61][62][63][64].   Figure 9 showed the obvious structural differences of three used dye molecules including MB, RhB and ST together with the top and side view of their space-filling models. The common feature of these three dyes was that they all contained large heterocyclic conjugated systems. The dimensions of the organic dyes are defined as x × y × z, wherein x, y and z represent the three edges of a cuboid that can just include the molecule. MB is one kind of planar molecules with x value of 16.3 Å, y value of 7.9 Å and z value of 4.0 Å. RhB is cationic with a triangle structure and contains a carboxylic group, wherein the values of x, y, z are 17.9 Å, 9.8 Å and 4.3 Å, respectively. In addition, ST demonstrates a T-type structure with x value of 16.1 Å, y value of 9.6 Å and z value of 4.3 Å [49][50][51][52][53][54][55]. Combined with the above results, schematic illustration of the preparation process of MXene-SO3H and the selfassembled modes in MXene-SO3H/dye composite Langmuir films were shown in Figure 10. At first, the targeted amphiphilic MXene-SO3H were prepared by intercalation of Na + ions and the modification of sulfanilic acid diazonium salts. Then the MXene-SO3H/dye composites Langmuir films and interfacial self-assembled behaviors at the air-water interface have been characterized and investigated via π-A isotherms, AFM and FT-IR techniques. Depending on different molecular structures in three used model dyes, H-and/or J-aggregates could be formed in different MXene-SO3H composite LB films, demonstrating reasonable and precise regulation of composite films via design of suitable components in organized nanostructures [56][57][58][59][60][61][62][63][64].  In SERS experiment, the performed transfer layers number X (25, 50, 75 and 100) of MXene-SO3H/MB was changed for getting the optimum film substrate and the corresponding substrates are defined as MXene-SO3H/MB-X in the following discussion. Figure 11a indicated the Raman intensity of 10 -3 M R6G solution absorbed in the MXene-SO3H/MB film substrate with different layers. As we could see, the SERS activity of the different substrates climbed up and then declined along with the increase of layers. This result showed that MXene-SO3H/MB-50 substrate had the best SERS enhancement effect. Next, we conducted direct Raman measurements for the optimum substrate MXene-SO3H/MB-50 containing MB dyes. In order to calculate substrate -to-substrate reproducibility, SERS spectra of MB were obtained on 36 random locations from 6 different MXene-SO3H/MB-50 substrates, as shown in Figure 11b. Moreover, the location of MB characteristic peaks was consistent with previous experiments. In addition, homogeneity of Raman spectra on the whole area is very important when considering the practical application of SERS substrate. To obtain a statistically significant result, the peak area at the 1396 and 1629 cm −1 band of MB molecules with all 36 spectra in Raman mapping were shown in Figure 11c,e and the corresponding bar plot of intensity versus detection points were shown in Figure 11d,f. Almost same peak height and area were observed at all 36 points with the relative standard deviation (RSD) of Raman intensity, which were computed to be 5.3% for 1396 cm -1 and 4.9% for 1629 cm −1 , indicating the good uniformity of MXene-SO3H/MB-50 substrate in large area. In addition, the Raman intensity with observed was excited with a 532 nm laser in the whole experiment. In SERS experiment, the performed transfer layers number X (25, 50, 75 and 100) of MXene-SO3H/MB was changed for getting the optimum film substrate and the corresponding substrates are defined as MXene-SO3H/MB-X in the following discussion. Figure 11a indicated the Raman intensity of 10 −3 M R6G solution absorbed in the MXene-SO3H/MB film substrate with different layers. As we could see, the SERS activity of the different substrates climbed up and then declined along with the increase of layers. This result showed that MXene-SO3H/MB-50 substrate had the best SERS enhancement effect. Next, we conducted direct Raman measurements for the optimum substrate MXene-SO3H/MB-50 containing MB dyes. In order to calculate substrate -to-substrate reproducibility, SERS spectra of MB were obtained on 36 random locations from 6 different MXene-SO3H/MB-50 substrates, as shown in Figure 11b. Moreover, the location of MB characteristic peaks was consistent with previous experiments. In addition, homogeneity of Raman spectra on the whole area is very important when considering the practical application of SERS substrate. To obtain a statistically significant result, the peak area at the 1396 and 1629 cm −1 band of MB molecules with all 36 spectra in Raman mapping were shown in Figure 11c,e and the corresponding bar plot of intensity versus detection points were shown in Figure 11d,f. Almost same peak height and area were observed at all 36 points with the relative standard deviation (RSD) of Raman intensity, which were computed to be 5.3% for 1396 cm −1 and 4.9% for 1629 cm −1 , indicating the good uniformity of MXene-SO3H/MB-50 substrate in large area. In addition, the Raman intensity with observed was excited with a 532 nm laser in the whole experiment.   The Raman enhancement factor (EF) of MXene-SO3H/MB-50 substrate is calculated by use the 10 −3 R6G solution. The EF can be calculated by the following formula [65]: where Isurf and Ibulk represent the area of the same Raman vibrational band in the SERS substrate and the bulk sample, respectively. Nsurf is R6G molecules number excited by Raman scattering light on the obtained MXene-SO3H/MB-50 film substrate and Nbulk is the number of R6G molecules excited by Raman scattering light on the pure silicon. Figure 12 showed the R6G spectra intensity comparison result of the MXene-SO3H/MB-50 substrate with that the of the non-SERS substrate. Here, the same Raman vibrational band centered in 1363 cm −1 was chosen to compute the value of Isurf and Ibulk. The value of Isurf (~8.97 × 10 4 ) was got when the 10 µL 10 −3 R6G solution dropped on the 1 × 1 cm 2 MXene-SO3H/MB-50 substrate. Similarly, the Ibulk was calculated to be~8.57 × 10 3 . Therefore, the ratio of Isurf/Ibulk was 10.5 (8.97 × 10 4 /8.57 × 10 3 ). For R6G sample, the calculation formula of Nbulk is as follows [66]: where Slaser is the area of the laser spot (1 µm in diameter), d is penetration depth (about 10 µm), ρ (0.79 g/cm 3 ) is the density of solid R6G, NA (6.022 × 10 23 ) is the Avogadros number and M (479.01 g/mol) shows the molar mass of R6G. Therefore the value of Nbulk is 0.77 × 10 10 . Next, we assumed R6G molecules on the surface of MXene-SO3H/MB-50 film substrate to calculate Nsurf. The project area with one R6G molecule was about 2.0 nm 2 [67,68]. In addition, considering the surface morphology of MXene-SO3H/MB-50 substrate, we could calculate the surface area of the laser irradiated area, which was about 0.8 µm 2 , so the value of Nsurf was 4 × 10 5 . According to Formulas (2), Nbulk/Nsurf had a value of 0.19 × 10 5 (0.77 × 10 10 /4 × 10 5 ). Therefore, the present obtained MXene-SO3H/MB-50 substrate exhibited a EF of~1.995 × 10 5 according to Formulas (1 [33]. Compared to the above works, present prepared composite films via self-assembled LB approach showed good SERS performance with simple and organized process, demonstrating potential applications in SERS detections and composite materials [69][70][71][72][73][74][75][76][77].

Conclusions
In summary, we have presented the successful preparation of composite Langmuir films via chemically-modified MXene-SO3H particles and model dyes at the air-water interface and investigated their interfacial self-assembly as well as aggregation behaviors. The compositions of MB and RhB molecules with MXene-SO3H particles induced dye molecules to form H-and J-aggregation, while ST molecules mainly formed H-aggregation on MXene-SO3H surface. Comparing to MB and ST, the obtained RhB composite films demonstrated denser and formed large aggregation clusters. In addition, for potential surface-enhanced Raman scattering application, the MXene-SO3H/MB-50 film substrate exhibited a reasonable EF (1.995 × 10 5 ), low RSD (<16%), good uniformity and repeatability. The present research work could provide a new clue that the suitable chemical modification and self-assembly of 2D inorganic materials (MXene, graphene oxide, et al.) could form organized composite nanostructures as new biosensor substrate and probe matrix with simple preparation process.