High‐Resolution Mapping Nanoscale Hydrophobicity for Fine Structures and Dynamic Evolution of Nanomaterial Surface

Nanomaterial hydrophobicity plays a critical role in interfacial phenomena ranging from biological toxicity to chemical reactions. However, it is difficult to figure out the high‐resolution surface hydrophobicity at the nanoscale. Herein, a chemical force microscopy is demonstrated to profile in situ hydrophobicity images with the nanoscale resolution, exhibiting order‐of‐magnitude gain than the traditional methods. This method is utilized to rapidly recognize the spatial fine structure hydrophobicity on Au, graphite, mica, and graphene oxides (GO), enabling the recognition of complicated substances and structures. It is found that the hydrophobicity of GO is opaque and is independent of stacking thickness, which is entirely different from the original graphene. Especially, the regions of wrinkles/edges are first proved to be generally less attractive to the hydrophobic probe than flat areas. This method is used to observe the dynamic evolution of GO hydrophobicity in different aqueous conditions, and is capable of detecting local oxidation variation during interfacial reactions.


Introduction
Over the past decades, numerous types of nanomaterials have emerged and attracted intensive attention due to their outstanding properties in energy science, [1] environmental chemistry, [2] and biology. [3]nderstanding the surface hydrophobicity of nanomaterials plays a significant role in predicting their chemical reactions and achieving nanomaterial-enabled applications. [4]For example, recent researches pointed out that hydrophilic nanodots possessed efficient catalytic activity in transforming CO 2 into methanol, [5] and the surface hydrophobicity determines the nanomaterial cellular uptake, transport, and fate. [6]However, in past efforts, the surface hydrophobicity of nanomaterials used to be simply recognized as homogeneous and its measurement heavily relied on macroscale wettability method. [5]The hydrophobic characteristics in spatial heterogeneity and dynamic processes have been limited until now.Thus, the development of nanomaterials is posing novel challenges to understand the surface hydrophobicity at the nanomaterial solid-water interface.
Apart from the traditional water contact angle (WCA) method, [7] only a few methods are available for determining the hydrophobicity of nanomaterials, including surface adsorption assays [8] and relative affinity for reference phases. [9]owever, these methods fail to figure out hydrophobic heterogeneity at the microscale and nanoscale.To recognize the surface hydrophobicity mapping, some strategies were proposed including the optical microscopy and molecule probe techniques. [10]10b,10c] These techniques still suffer from their inherent shortcomings, including limited spatial resolution and inapplicability to all nanomaterials.Recent developments of atomic force microscopy (AFM)-based force spectroscopy were applied to characterize the interfacial hydrophobicity qualitatively. [11]Utilizing modified hydrophobic tips, AFM-based force spectroscopy delivers the potential to detect hydrophobic interaction of sample surfaces at the microscale level. [12]Recently, a newly developed 3D-AFM microscope enables atomic-scale mapping of hydrophobic layers on graphene. [13]But its sluggish speed of probe movement and complicated operation limit the accessibility of high-resolution mapping for nanoscale hydrophobicity.To further understand the fine nanostructures and dynamic evolution of nanomaterial surface, an advanced hydrophobicity mapping technique with higher resolution and faster scan speed is required.
Herein, we developed a facile strategy that combined chemical force spectroscopy (CFM) and PeakForce Tapping mode to achieve high-speed and high-resolution 3D mapping on graphene oxide (GO) hydrophobicity.GO single-layer sheet is one the most potential nanomaterials with outstanding properties and is the focus of this study. [14]Studying surface hydrophobicity on GO provides a framework for understanding hydrophobic structures of nanomaterials.Through hydrophobicity recognition within 20 nm enabled by Fast PeakForce Tapping mode and CFM, we elucidate that the hydrophobicity of GO was opaque and was independent of stacking thickness due to the short-range chemical bonding from hydration layers.Especially, the wrinkles/edges regions on GO nanosheets were directly observed less attractive to the hydrophobic probe than the flat regions for the first time, and the dynamic variations on GO hydrophobicity were systematically detected under in situ aqueous conditions, including pH, ionic strength, and humic acid.Our efforts made a breakthrough in developing a visualized technology for in situ nanoscale hydrophobicity, and provided indepth insights into spatial structures and the dynamic evolution of nanomaterial surface hydrophobicity.

High-Resolution Mapping of Hydrophobicity Based on Force Spectroscopy
12a] Such a hydrophobic CH 3 (CH 2 ) 7 -terminated monolayer on the AFM tip could also screen intermolecular attractions caused by hydrogen bonding or electrostatic forces.The surface of AFM tip remained smooth after modification (Figure S1, Supporting Information) and the C/Si elemental mass ratio of the modified AFM tip increased by 39.6% (Table S1 and Figure S2, Supporting Information), indicating that OTS molecules were successfully deposited onto the tip surface.Further WCA measurements also showed that the modified surface possessed extremely hydrophobic WCA of 95.4 AE 0.5º (Figure S3, Supporting Information), which was much larger than that of the unmodified surface (72.6 AE 4.9º).
A schematic illustration of measuring hydrophobic interaction is shown in Figure 1a.12a,15] The functions between surface hydrophobicity and adhesion force are deduced in the Supporting Information, which is based on the Young equation and the Johnson-Kendall-Roberts model.To verify the relationship between the adhesion forces and hydrophobicity of the sample surface, AFM adhesion mapping was carried out on model substrates of varying hydrophobicity.Model substrates were gold samples modified with mixed thiol at different ratio (CH 3 (CH 2 ) 10 SH and HO(CH 2 ) 11 SH in ethanol).The WCA raised with increased content of CH 3 (CH 2 ) 10 SH due to its hydrophobicity (Figure S4, Supporting Information).The magnitude of the adhesion force exhibits an increasing trend with the rise in WCA of substrate (Figure 1b), which can be attributed to the stronger hydrophobic interactions at the hydrophobic interface.This observation suggests that a higher contact angle corresponds to a greater hydrophobic force, resulting in an enhanced adhesion force between the probe and the substrate.An almost linear relationship can be drawn by comparing adhesion forces with the cosine value of WCA (Figure 1b).To improve the mapping speed and resolution, the Bruker PeakForce Tapping technique was adopted to drive the AFM tip, which could finish a single approach/retract cycle at every pixel location at a rate of %2,000 s À1 .Figure 1c,d shows the adhesion mapping for hydrophobic sites of model substrate SF1 (The height image was shown in Figure S5, Supporting Information).It was noted that the pixel resolution reached up to less than 20 nm, achieving the highresolution detection for surface hydrophobicity in the literature.

High-Resolution Visualized Hydrophobicity Images at the GO-Water Interface
Single-layer GO was deposited on the newly cleaved mica surface for analytical purposes.The lateral size of GO was %3 Â 3 μm (Figure 2a), and there were wrinkles across the GO sheet.The corresponding surface hydrophobicity of GO is shown in Figure 2b, enabling the detection of nanoscale spatial variations of GO with high resolution.Figure 2b can be divided into three parts according to the values of adhesion forces.The regions of mica showed higher adhesion forces with OTS-modified tip than that measured flat area of GO sheet.7b,16] Since hydrophobicity transparency was related to extreme thinness and was dominated by short-range chemical bonding, [7c] the hydrophobicity opacity of GO was inferred to be caused by the shielding effect of thick hydration layer.The hydrophobicity image of GO was observed to be not homogeneous, reflecting the difference of sp 2 and sp 3 regions and the fluctuations from water molecules at the GO interface.To the best of our knowledge, it was the first time that heterogeneous hydrophobic interactions of hydrated GO surface were directly measured in situ.Wrinkles were unexpectedly observed to be less attractive to the hydrophobic probe through comparison between flat regions and wrinkles of GO.As illustrated by the cross-sectional adhesion force of dotted line in Figure 2b (Figure 2e-g shows the representative force-separation curves in Figure 2b), the adhesion forces between OTS molecules and wrinkles of GO were less than 100 pN.The adhesion forces showed an obvious trend from decline to rise while crossing the wrinkle of GO.According to the classical GO model with both sp 2 -hybridized areas and sp 3 -hybridized areas, a high edge-to-area ratio (high surface energy) from wrinkle edges could strengthen the affinity with water molecules.In addition, the structural effect of wrinkles might expose more epoxy and hydroxyl groups and thus reduce contact with OTS molecules.And the change in Young's modulus caused by wrinkles and the alteration in contact area with the probe may contribute to the decrease in adhesion observed in our study.Figure 2c depicts the histogram and Gaussian fitting of adhesion forces in overall adhesion mapping, in which the adhesion forces of OTS-modified AFM tip on GO and mica were %0.19 and 0.27 nN, respectively.It suggested that GO was extremely hydrophilic (the calculated cosine value of WCA was 0.96) by referring to the linear relationship, as shown in Figure 1b.To vividly display the relationship between the morphology and hydrophobicity of GO, we drew a 3D height topography image with adhesion mapping, as skin in Figure 2d.This made it easy to extract the whole information of GO from a single image plane simultaneously.
7c] To classify this dispute, several single-layer GO sheets were stacked layer-by-layer (Figure 3a).It could also be seen that the wrinkled area was relatively less attractive to the hydrophobic probe than the flat area (Figure 3b).And the histogram of adhesion forces in Figure 3c can roughly be divided into two Gaussian distribution fitting, which corresponds to the hydrophobic interactions of GO and mica (GO: %0.07 nN, mica: %0.16 nN), respectively.It suggested that hydrophobicity of GO was different from that of mica substrate, and thus exhibited opaque.7c,15] However, adhesion mapping in Figure 3b shows that there was no difference in hydrophobic interaction between stacked and nonstacked areas.Hence, the hydrophobicity of GO was not dependent on the thickness, unlike the properties of graphene.This might be because the hydration layers on the surface of the oxidized graphene made each GO sheet an insulating and independent individual in terms of hydrophobicity.Figure 3d vividly exhibits the relationship between surface structure and hydrophobicity of GO in a 3D form.More images of hydrophobic interactions for GO sheets are shown in Figure S6, Supporting Information.From these images, we could also confirm that the hydrophobic sites of GO were not affected by layer-stacking.The hydrophobicity of each GO sheet was confirmed to be different: GO sheets in Figure 2b and Figure 3b were more hydrophilic than mica, yet some GO sheets were less hydrophilic than mica, as shown in Figure S7, Supporting Information.7c] To exclude the above possibility, we selected highly oriented pyrolytic graphite (HOPG) as substrate to substitute mica because the hydrophobicity of HOPG was not dominated by hydrogen bonding with water.Figure 3 and Figure S8, Supporting Information show the hydrophobic images of a GO monolayer that was stacked on the HOPG.GO was extremely hydrophilic in contrast to HOPG.Hydrophobic attractive forces between OTS molecules and HOPG were fitted as 4.4 nN by Gaussian distribution (Figure 3g), which was much larger than the adhesion forces between GO and OTS molecules (0.36 nN).This result further confirmed that GO was opaque to hydrophobicity no matter what the substrate it was.In addition, the distinct difference in hydrophobic interaction between GO and HOPG in Figure 3f also inspired us the accessibility to recognize the sp 2 and sp 3 sites through this AFM-based method.
As aforementioned, interfacial aqueous layers were considered as a critical role in hydrophobic interaction sites for hydrated GO.We detected the existence of interfacial aqueous layers by comparing the GO height in air (Figure 4a) and in water (Figure 4b).The height of GO in the air was around 1.3 nm.After 10 min equilibrium in water, the height of this GO sheet in water increased to %1.7 nm (Figure 4c).Since the hydration layers cannot be directly detected in the height images, [17] the increase in GO height indicated the intercalation of an interfacial aqueous layer between GO flakes and the mica substrate. [18]The thickness of the primary interfacial aqueous layer is approximately 3 angstroms, [19] which corresponds to the size of one water molecule.The presence of height fluctuations in the interfacial aqueous layers indicates that different sites on GO demonstrate different affinities for water.This observation provides evidence of the varying hydration abilities across different sites on the GO layers.These interesting phenomena suggested that heterogeneous hydration layers would cover GO sheets and influence their surface hydrophobicity.

Detecting Dynamic Variations of GO Hydrophobicity in Water Chemistry
In addition to the nature of the interacting GO surfaces, hydrophobic interaction depends on water chemistry.To detect the dynamic variations of GO hydrophobicity in water chemistry, the hydrophobicity was measured via OTS modified AFM probe in varied aqueous conditions, including varying pH (3, 6, 11), ionic strength (10 mM NaCl and 50 mM NaCl), and humic acid (50 and 500 mg L À1 ). Figure 5a-c depicts the in situ variation of hydrophobicity of GO in the solution of pH 3, 6, and 11.Notably, the hydrophobic interaction on GO declined with the increase in pH value, indicating that the hydrophilicity of GO gradually improved.According to previous studies, alkaline solution conditions would boost the dissociation of oxygen functional groups on the GO surface, thereby thickening the hydration layer (Figure S9a, Supporting Information). [20]The thickening hydration layers could improve the hydrophilicity of GO.Generally, ionic strength was considered as a key impact factor to alter the environmental interfacial behaviors of GO.Herein, we studied the influence of ionic strength on the hydrophobicity of GO. Figure 5d-f shows the hydrophobic interaction mapping of GO in DI water, 10 mM NaCl, and 50 mM NaCl, respectively.Interestingly, ionic strength had little influence on hydrophobicity of GO according to the variations between Figure 5d-f.While measuring in an aqueous environment, the potential changes in tip radius during AFM measurements can introduce uncertainties in the measured adhesion forces.However, in our study, we observed no significant increasing trend in the adhesion force, indicating that the tip radius remained relatively constant during the probing process.This finding strengthens the reliability of our measurements and underscores the importance of monitoring and accounting for any variations in tip radius for accurate and consistent AFM measurements.
Humic acid is an important class of natural organic matter in aquatic environments.Hence, the influence of humic acid on the hydrophobicity of GO is of significance to the environmental behaviors of GO.The influence of humic acid on GO hydrophobicity is shown in Figure 5g-i, which corresponded to GO in DI water, 50 mg L À1 humic acid, and 500 mg L À1 humic acid, respectively.By contrasting Figure 5g,h, it was observed that the hydrophobic interaction on the surface significantly increased (Note: the legend value in Figure 5g,h was different).When the concentration of humic acid increased to 500 mg L À1 , hydrophobic interaction further improved and some deposition was observed on the surface.Because humic acid contained both hydrophilic and hydrophobic tails similar to surfactant, its hydrophobic tails could be extruded into the water after adsorption on hydrophilic regions, leading to the increase of hydrophobicity (Figure S9c, Supporting Information).

Perspective for Single Molecular Force Estimation
It should be noted that the measured force was the overall force between GO and OTS molecules on AFM tips, which was dominated by the size of the AFM tip.To avoid the effect of tip size, an estimation of hydrophobic interaction of a single OTS molecule on GO was given below where F OTS is the molecular interaction of a single OTS on GO, F ad is the adhesion force between the OTS-modified AFM tip and GO, N OTS is the number of OTS molecules in tip-GO contact area, S OTS is a cross-sectional area per OTS molecule, and S ad is the tip-GO contact area.The value of S OTS could refer to an average area occupied by a terminal methyl group in the silylized layer, which is 0.40 square nm. [21]S ad could be calculated by the equation, S ad = πR ad 2 , where R ad is the contact radius.R ad could be estimated by the Johnson-Kendall-Roberts theory for mechanical contacts.
where w is the adhesion work per unit area, R is the radius of the AFM tip, and K is the reduced modulus related to the Young's moduli of silicon AFM tip and GO.22a,23] Using Equation (1-3), the molecular interaction of a single OTS on GO could be written as The radius of the AFM tip was 20 nm.Thus, Equation (4) could give an estimate for F OTS .According to the above CFM results, the measured F ad varied from 0.07 to 0.88 nN due to nonuniform hydrophobicity of GO sheets.Hence, F OTS was calculated as 2.3-5.4 pN after normalization.However, considering that the uncertainties of both K, S OTS , R, and growth density of OTS monolayers, the calculated result for F OTS was approximate.Considering these uncertainties for F OTS , it could be concluded that the value of a single OTS molecule force on GO was approximately several pico-Newton forces.This approach could provide a quantitative determination for studying the molecular-level interface behaviors of GO in aqueous conditions.

Conclusion
In summary, chemical force microscopy based on PeakForce Tapping characterization has been utilized to map the hydrophobic interaction sites of hydrated GO with high speed and high resolution.AFM tips were modified with hydrophobic CH 3 (CH 2 ) 7 -terminated groups via chemical vapor deposition of (octyl)-trimethoxysilane, and the linear relationship between surface hydrophobicity and adhesion forces was verified by AFM in water.Using this approach, we investigated hydrophobic interaction sites of GO sheets at the nanoscale.It was found that the hydrophobicity of GO was opaque, and it was independent of stacking thickness.Particularly, the hydrophobicity on GO basal plane was heterogeneous, and the regions of wrinkles/edges were observed to be less attractive to the hydrophobic probe than flat regions.The results indicated that alkaline solution conditions could improve hydrophilicity of GO by boosting the dissociation of functional groups.Adsorption of humic acid changed the hydrophobicity state.However, ionic strength had little influence on hydrophobic sites of GO.In addition, the value of single OTS molecule force on GO was estimated to be 2.3-5.4 pN after normalization.Our efforts were of significance to design advanced functions and predict environmental fates of GO.Future efforts could apply this demonstrated method to provide dynamic insights into the spatial and temporal variations for hydrophobicity of nanomaterials.
Modification of AFM Tips and Model Substrates: To gain a hydrophobic surface, AFM tips were modified by (octyl)-trimethoxysilane via vapor deposition.Prior to modification, AFM tips were cleaned and activated to generate -OH groups by O 2 plasma in the plasma cleaning machine (YZD08-2C, Saot Optoelectronic Technology).5 μL of (octyl)trimethoxysilane (OTS) was vaporized by heating and then deposited on AFM tips in a 500 mL vacuum container at 110 °C for 24 h.(Octyl)trimethoxysilane formed as self-assembly layers on AFM tips via Si-O-Si bonds. [25]Finally, the modified AFM tips were blown to remove unreacted silane with high-purity nitrogen.Model substrates of varying hydrophobicity were modified with thiol, and then used to calibrate hydrophobic force spectroscopy of AFM tips.First, Au-coated silicon wafers were cleaned with ethanol and O 2 plasma, respectively.Then these wafers were immersed in 1 mM of mixed thiol solution (ethanol used as solvent) for 24 h at room temperature.Thiols formed self-assembly layers via Au/-SH bonds. [26]The ratio of CH 3 (CH 2 ) 10 SH and HO(CH 2 ) 11 SH in mixed thiol solution was set as 1:0, 0.8:0.2,0.6:0.4,0.4:0.6,0.2:0.8, and 0:1, respectively, which led to a series of model substrates of varying hydrophobicity (denoted as SF1, SF2, SF3, SF4, SF5, and SF6).Finally, these model substrates were washed with ethanol and blown dry with high-purity nitrogen.
Characterizations: To fix single-layer GO, 40 μL of diluted 4.0 μg g À1 GO aqueous dispersion was dropped on newly cleaved mica discs (proper concentration is critical to forming single-layer sheets by avoiding stacking of GO).Then GO was dried at room temperature to make single-layer GO bind to the mica surface via intermolecular forces.To enhance the binding of GO to the hydrophobic HOPG substrate after the natural drying process, an additional step of freeze-drying was implemented in the preparation process.The samples were frozen in a cold trap set at À60 °C.Subsequently, the frozen samples were placed inside a freeze dryer, where we created a vacuum within the chamber while maintaining a temperature of À60 ºC for a duration of 24 h.This process facilitated the removal of frozen water molecules as vapor, effectively eliminating any residual moisture between the GO and mica surfaces.The morphology and thickness of GO were characterized in both air and water by AFM (Dimension Icon, Bruker).Surface hydrophobicity was measured by the optical surface analyzer (OSA200, Ningbo NB Scientific Instruments) via the water contact angle method.All the water contact angles were measured via 2 μL of deionized water droplets at least three tests.A scanning electron microscope (SU8010, Hitachi) with energy dispersive spectrometer was utilized to analyze the modified AFM tips.
Hydrophobic Sites Mapping by CFM: Chemical force spectroscopy determined hydrophobic interaction sites based on adhesion forces in water using modified hydrophobic AFM tips.AFM system was controlled by Bruker Nanoscope 9.3 software.The sensitivity deflection and force constant of the modified AFM tip were calibrated by the thermal tuning method in accordance with the Bruker Guide. [27]The force constant was around 1.1-1.3nN nm À1 in water.Adhesion forces between the modified AFM tips and substrate surface were measured by PeakForce Tapping at a load of 5 nN and a scan rate of 1 kHz.The ramp amplitude was set as 100 nm.The adhesion force (F) was calculated by F=kΔz, where k is the force constant and Δz is the tip deflection.On average, 128 Â 128 force curves were measured on each PeakForce Tapping for each water chemistry (including pH of 3, 6, 11, or ionic strength of 10 mM NaCl and 50 mM NaCl, or introducing 50 and 500 mg L À1 humic acid).Quality control: to maintain consistency in the measurement of hydrophobicity, we have specifically used the same tip for the same batch of experiments.When comparing different plots, the presence of mica has been introduced as a reference system.The adhesion force at the mica shows a narrow fluctuation range of 0.16-0.29 nN, with a mere 0.13 nN variability.This systematic error caused by tip chemistry change does not affect our intuitive observation and comparison.Analysis of AFM data was carried out using Bruker Nanoscope Analysis 1.9 software.The average and standard error of adhesion forces were calculated from all the force curves.A total of 16 384 pixels (128 Â 128) of adhesion forces in the image were collected and analyzed by the histogram, which could help understand the overall hydrophobicity of samples (Notice: The imaging resolution is primarily influenced by the radius of the AFM tip.Adjusting the sampling frequency with a sharper tip can improve resolution.However, the thermal drift of the piezoelectric scanner and the imaging time also impact resolution.Higher sampling frequency for higher resolution requires more time.Prolonged imaging time can result in noticeable image distortion due to thermal drift.Additionally, when using a sharper AFM tip for adhesion force measurements, the measured forces were extremely small, making it challenging to detect significant differences).

Figure 1 .
Figure 1.a) Schematic of detecting the hydrophobic sites by AFM.b) Adhesion forces measured related to the cosine of water contact angle of substrates (adhesion force measurements: 25 different locations across three samples on each substrate, water contact angle measurements: three different locations across three samples on each substrate).c) Two-dimensional mapping of adhesion forces on hydrophobic substrate SF1 in water, and pix number is 128 Â 128.d) The histogram of adhesion forces in mapping.

Figure 2 .
Figure 2. a) Height image of GO in water and its cross-sectional height of dotted line.b) Adhesion force mapping of GO in water and its cross-sectional adhesion force of dotted line.c) The histogram and Gaussian fitting of adhesion forces.d) 3D topography of height images with adhesion mapping as skin.e-g) Representative force-separation curves in the area of 1, 2, and 3, respectively in Figure 3b.Blue line: approaching, red line: retracting.

Figure 3 .
Figure 3. a) Height image of stacked GO on mica in water, b) its adhesion force mapping, c) its histogram and Gaussian fitting of adhesion forces, andd) its 3D height topography image with adhesion mapping as skin.e) Height image of GO on highly oriented pyrolytic graphite (HOPG) in water, f ) its adhesion force mapping, g) its histogram and Gaussian fitting of adhesion forces, and h) its 3D height topography image with adhesion mapping as skin.

Figure 4 .
Figure 4. a) AFM height image of GO measured in air.b) AFM height image of GO measured in DI water after 15 min equilibrium.c) The cross-sectional height of the dotted line.

Figure 5 .
Figure 5.The dynamic variations of GO hydrophobicity in aqueous solution.GO in the solution of a) pH 3, b) pH 6, and c) and pH 11. d) GO in DI water, e) 10 mM NaCl, and f ) 50 mM NaCl.g) GO without humic acid, h) 50 mg L À1 humic acid, and i) 500 mg L À1 humic acid.