Laser-Induced Graphitization of Polydopamine on Titania Nanotubes

Since the discovery of laser-induced graphite/graphene, there has been a notable surge of scientific interest in advancing diverse methodologies for their synthesis and applications. This study focuses on the utilization of a pulsed Nd:YAG laser to achieve graphitization of polydopamine (PDA) deposited on the surface of titania nanotubes. The partial graphitization is corroborated through Raman and XPS spectroscopies and supported by water contact angle, nanomechanical, and electrochemical measurements. Reactive molecular dynamics simulations confirm the possibility of graphitization in the nanosecond time scale with the evolution of NH3, H2O, and CO2 gases. A thorough exploration of the lasing parameter space (wavelength, pulse energy, and number of pulses) was conducted with the aim of improving either electrochemical activity or photocurrent generation. Whereas the 532 nm laser pulses interacted mostly with the PDA coating, the 365 nm pulses were absorbed by both PDA and the substrate nanotubes, leading to a higher graphitization degree. The majority of the photocurrent and quantum efficiency enhancement is observed in the visible light between 400 and 550 nm. The proposed composite is applied as a photoelectrochemical (PEC) sensor of serotonin in nanomolar concentrations. Because of the suppressed recombination and facilitated charge transfer caused by the laser graphitization, the proposed composite exhibits significantly enhanced PEC performance. In the sensing application, it showed superior sensitivity and a limit of detection competitive with nonprecious metal materials


Materials and methods -supplementary
Reagents 10 X Tris Buffer was obtained from Santa Cruz Biotechnology and diluted 10 times prior to the measurements, dopamine hydrochloride was purchased from Sigma-Aldrich and the deionized water used in the experiment was provided by a Hydrolab HLP-5p system.Titanium plate (99.7% purity) was obtained from Strem Chemicals.Na 2 SO 4 , acetone, ethanol, diethylene glycol, HCl, NaOH, K 3 [Fe(CN) 6 ], K 4 [Fe(CN) 6 ], and NH 4 F were provided by Chempur.

Synthesis of titania nanotubes (TNTs)
The protocols for titania nanotube synthesis were reported previously elsewhere. 27,28Briefly, prior to anodization, the titanium plate was ultrasonically cleaned in acetone, ethanol and subsequently water.
Single-step anodization of the titanium plate was performed in an electrolyte containing diethylene glycol / water / HF (69 / 5 / 1 vol.%) and 9.4 mM NH 4 F. The anodization voltage was 30 V or 40 V (as indicated in the text), the polarization time was 2 h and the temperature was 40 °C.After anodization, the TNTs were rinsed with ethanol and dried in an air stream.Finally, the plates were calcined in a tube furnace (Nabertherm) at 450 °C for 2 hours.

X-Ray Photoelectron Spectroscopy
X-Ray Photoelectron Spectroscopy (XPS) studies were carried out on an Escalab 250Xi multispectroscope (ThermoFisher Scientific) operating with an AlKα X-ray source.The spot size was 650 µm and the pass energy through the hemispherical analyzer was 20 eV.Throughout the measurement, the samples were flooded with low-energy electrons and low-energy Ar + ions to assure charge compensation, with a final peak calibration at adventitious carbon C 1s (284.8 eV).The obtained spectra were analyzed and deconvoluted using the Avantage v5.9921 software (ThermoFisher Scientific).

Diffuse reflectance UV-vis spectroscopy
Diffuse reflectance spectra were collected with a PerkinElmer Lambda 35 (PerkinElmer, Waltham, MA, USA) dual-beam spectrophotometer in the range of 200-1000 nm with a scanning speed of 60 nm/min.The Kubelka-Munk functions F(R) were calculated on the reflectance R data according to the Equation 2 1 :

Raman spectroscopy
Raman spectroscopy measurements were conducted using a confocal micro-Raman spectrometer (InVia, Renishaw, Wotton-under-Edge, Gloucestershire, UK) with an argon-ion laser source emitting at 514 nm and operating with a power of 10 mW.Raw Raman data were normalized with respect to the largest anatase band (A band) and the fluorescence was subtracted to highlight the G/D band ratios.The deconvolution process involved fitting two straight lines so that the fluorescence background signal beneath the D and G bands is reduced to zero.

Density functional theory calculation of the lgPDA electronic structure
Density functional theory (DFT) computations were performed using the Perdew-Burke-Ernzerhof (PBE) functional within the generalized gradient approximation (GGA) implemented in the ATK package 2 .
The Linear Combination of Atomic Orbitals (LCAO) 3 method with medium ATK basis set and Fritz-Haber Institute (FHI) pseudopotentials were applied.
The lgPDA structure was taken from the final frame of the ReaxFF simulation and all the gases were removed.Then, the resulting lgPDA sheet was DFT optimized with relaxed parameters and symmetry of the unit cell with 0.05 eV/Ang threshold.
Density of states and bandstructures were calculated with a 7×7×1 A 3 k-point mesh and a density mesh cutoff of 125 Ha.The DOS spectra of isolated PDA tetramer was computed for their corresponding optimized structures with multipole boundary conditions (single k-point).

Results -supplementary
CV curves for dopamine electropolymerization on adjacent titania nanotubes

Figure S1 .
Figure S1.50 cycles of cyclic voltammetry with 20 mV/s scan speed registered during dopamine electropolymerization in the Ar-purged solution containing 5 mM dopamine, 1 X tris and 0.5 M Na 2 SO 4 .Arrows indicate the direction of current changes during cycling.

Figure S2 .
Figure S2.SEM images of the modified loosely spaced nanotubes; a-b) images of the pristine TNTs; c-d) images of the TNT_PDA before graphitization; e-f) images of the TNT_lgPDA with 532 nm exposure.Two contrasts of the same images are given for clearer representation.

Figure S3 .
Figure S3.Optimization of the anodization and electropolymerization parameters prior graphitization; a) CVs of PDA electropolymerization given for nanotubes synthesized with varying anodization time -the first cycle of 50 is shown; b) CVs in ferrocyanide solution registered after PDA electropolymerization with varying number of cycles at pH = 7.5; c) CVs in ferrocyanide solution registered after 50 cycles of PDA electropolymerization with varying pH; d-f) visible photocurrent density -potential dependencies of TNT_PDA depending on the anodization time (d), number of PDA electropolymerization cycles (e) and electropolymerization pH (f).The b), c), e) and f) curves are presented for TNT_PDA samples with nanotubes synthesized during 2h anodization.Electrolyte for ferrocyanide experiments contained 5 mM Fe(CN) 6 3-/ 5 mM Fe(CN) 6 4-).Supporting electrolyte was 0.5 M Na 2 SO 4 for every measurement.

Figure S4 .
Figure S4.SEM image of the TNT_PDA structure based on the adjacent nanotubes fabricated during 3hour anodization.Note the nanotubes morphology is lost after PDA deposition.

Figure S6 .
Figure S6.Details of the multiload nanoindentation experiments for pristine TNTs, TNT_PDA and TNT_lgPDA samples; a-c) load-displacement curves of the multiload procedure; d-f) deformation-time curves of the first indent; g-i) deformation-time curves of the second and fifth indents used for calculation of mechanical properties presented in the main text.

Figure S7 .
Figure S7.Statistics of the nanotube wall thickness of TNT, TNT_PDA, TNT_lgPDA samples calculated from the SEM images of the loosely spaced nanotubes in Figure S2 used to estimate the coating thickness given in Figure 2 in the main text.

Figure S9 .
Figure S9.Optimization of laser parameters for TNT_lgPDA modified using 365 nm wavelength.Top pane illustrates the effect of the laser fluence on a) Raman spectra, b) CV curves in dark (0.5 M Na 2 SO 4 ), c) CV curves in presence of ferrocyanides (5 mM Fe(CN) 6 3-/ 5 mM Fe(CN) 6 4-), d) corresponding EIS spectra, d) photocurrent -potential dependence in the visible part of the solar spectrum (0.5 M Na 2 SO 4 ).F30-F120 values correspond to the range of laser fluences between 30 mJ/cm 2 to 120 mJ/cm 2 .Bottom panel illustrates the effect of the number of the laser movement speed (inverse of the number of pulses) on f) Raman spectra, g) CV curves in dark, h) CV curves in presence of ferrocyanides, i) corresponding EIS spectra, j) photocurrent -potential dependence in the visible part of the solar spectrum.S25-S200 values correspond to the range of motorized table speed between 25 mm/min to 200 mm/min.

Figure S10 .
Figure S10.Electrochemical properties and Raman spectra of the TNT_lgPDA structures based on loosely spaced nanotubes; a-b) CV curves and EIS spectra obtained for different fluences of the 532 nm modification; c-d) CV curves and EIS spectra obtained for different fluences of the 365 nm modification; e) Raw Raman data for the 365 nm modification with different fluences; f) comparison of the Raman spectra obtained for the graphitized and pristine PDA after fluorescence subtraction.

Figure S12 .
Figure S12.IMPS Bode and Nyquist spectra of the pristine and modified TNTs recorded during +500 mV polarization and AC/DC light intensity equal to 0.8 mW / 1 mW at 424 nm wavelength.

Figure S13 .
Figure S13.IMVS Bode and Nyquist spectra of the pristine and modified TNTs recorded in open circuit and AC/DC light intensity equal to 0.2 mW / 1 mW at 373 nm wavelength.

Figure S14 .
Figure S14.Electronic structure of the graphitized PDA; a) lgPDA sheet taken from ReaxFF MD simulation, b) DFT-optimized geometry of the resulting sheet; c-d) surface bandstructure and density of states of the sheet; e) density of states of single DHI tetramer.