Laser Fragmentation‐Induced Defect‐Rich Cobalt Oxide Nanoparticles for Electrochemical Oxygen Evolution Reaction

Abstract Sub‐5 nm cobalt oxide nanoparticles are produced in a flowing water system by pulsed laser fragmentation in liquid (PLFL). Particle fragmentation from 8 nm to 4 nm occurs and is attributed to the oxidation process in water where oxidative species are present and the local temperature is rapidly elevated under laser irradiation. Significantly higher surface area, crystal phase transformation, and formation of structural defects (Co2+ defects and oxygen vacancies) through the PLFL process are evidenced by detailed structural characterizations by nitrogen physisorption, electron microscopy, synchrotron X‐ray diffraction, and X‐ray photoelectron spectroscopy. When employed as electrocatalysts for the oxygen evolution reaction under alkaline conditions, the fragmented cobalt oxides exhibit superior catalytic activity over pristine and nanocast cobalt oxides, delivering a current density of 10 mA cm−2 at 369 mV and a Tafel slope of 46 mV dec−1, which is attributed to a larger exposed active surface area, the formation of defects, and an increased charge transfer rate. The study provides an effective approach to engineering cobalt oxide nanostructures in a flowing water system, which shows great potential for sustainable production of active cobalt catalysts.


Electrochemical Measurements
Electrochemical measurements were performed in a three-electrode configuration using a rotating disc electrode (Model: AFMSRCE, PINE Research Instrumentation); a hydrogen reference electrode (HydroFlex, Gaskatel) and Pt wire were used as reference electrode and counter electrode respectively. KOH (1 M) solution was filled in a Teflon cell as electrolyte. Before the electrochemical measurement, argon was purged through the cell for 30 min to remove oxygen. The temperature of the electrolyte in the cell was kept at 25 o C using a water circulation system. Working electrodes were fabricated by depositing target materials on glass carbon (GC) electrodes (PINE, 5 mm diameter, 0.196 cm 2 area). Before depositing materials, a thorough surface clean was performed by polishing the GC electrodes with Al 2 O 3 suspension (5 and 0.25 μm, Allied High Tech Products, Inc.). Afterwards, 4.8 mg of powder sample was dispersed in a mixed solution of 0.75 mL of H 2 O, 0.25 mL of 2-propanol and 50 μL of Nafion (5% in a mixture of water and alcohol). Then the mixture solution was treated in sonication bath for 30 min to form a homogeneous ink. After that, 5.25 μL of catalyst ink was dropped onto the GC electrode and dried under light irradiation for 10 min. The catalyst loading was calculated to be 0.12 mg/cm 2 in all cases for GC electrodes.
The linear sweep voltammetry (LSV) curves were collected by sweeping the potential from 0.7 V to 1.7 V vs RHE with a scan rate of 10 mV/s. Cyclic voltammetry (CV) measurements were carried out in a potential range between 0.7 and 1.6 vs RHE with a scan rate of 50 mV/s. In all measurements, the IR drop was compensated at 85%, and a rotating disc electrode configuration was kept a rotation speed of 2000 rpm. To perform stability tests, an electrode was firstly fabricated by drop-casting the catalyst ink on the surface of pre-washed nickel foam (0.5×1 cm 2 ) with 3.5 M HCl solution in an ultrasound bath for 10 min. A catalyst loading was around 1 mg/cm 2 . Then, the potential was recorded on this catalyst@Ni foam electrode at a constant current of 5 mA over a period of 18 h electrolysis.

Characterization
Powder X-ray diffraction (XRD) patterns were collected at room temperature on a Stoe theta/theta diffractometer in Bragg-Brentano geometry (Cu Kα 1/2 radiation) with a secondary monochroamtor. Data were measured with a proportional counter working as a point detector. Transmission electron microscopy (TEM) images of samples were measured at 100 kV by an H-7100 electron microscope from Hitachi. High resolution TEM (HR-TEM) and scanning electron microscopy (SEM) images were taken on HF-2000 and Hitachi S-5500 microscopes, respectively. N 2 -sorption isotherms were measured using 3Flex Micrometrics at 77 K. Prior to the measurements, the samples were degassed at 150 °C for 10 h. Brunauer-Emmett-Teller (BET) surface areas were determined from the relative pressure range between 0.06 and 0.2. Xray photoelectron spectroscopy (XPS) measurements were carried out with a SPECS GmbH spectrometer with a hemispherical analyzer (PHOIBOS 150 1D-DLD). The monochromatized Al Kα X-ray source (E = 1486.6 eV) was operated at 100 W. An analyzer pass energy of 20 eV was applied for the narrow scans. The medium area mode was used as lens mode. The base pressure during the experiment in the analysis chamber was 5 x 10 -10 mbar. The binding energy scale was corrected for surface charging by use of the C 1s peak of contaminant carbon as reference at 284.5 eV. The Fourier transform infrared spectroscopy was collected on PerkinElmer UATR Two using a diamond crystal, the number of scan was 32 and the resolution was 4 cm -1 . Ultraviolet-visible (UV-vis) extinction spectroscopy was used to determine the light extinction properties of the initial cobalt oxide powders dispersed in water. Extinction measurements were performed in transmission by using an extinction-calibrated spectrometer (Evolution 201, Thermo Fisher Scientific) and a quartz glass cuvette with 10 mm beam path. All samples were prepared equally for the measurement. The mass concentration of the powders in water was identical (0.0033 vol.-%). Before each measurement, the metastable dispersions were treated with ultrasonic for 1 min.
Synchrotron diffraction data were collected at the high resolution powder diffraction beamline (P02.1) at PETRA III (DESY). Samples were filled into 0.5 mm glass capillaries and mounted on the instrument sample spinner. Data were collected at 60 keV (λ=0.207200 Å) with a Perkin Elmer XRD1621 area detector. Rietveld refinements were performed with the program package TOPAS V6, BRUKER AXS.     Figure S4a     As marked in the IR spectra, the sharp absorption bands at around 547 cm -1 and 655 cm -1 are associated with the stretching vibration modes of tetrahedrally and octahedrally coordinated Co-O bond, respectively. In Figure S6a, the peak at 1109 cm -1 belongs to vibrations of C-O. Two peaks at 1558 and 1418 cm -1 in CoO could be assigned to the ethyl group from ethanol reduction ( Figure S6b). The absorption peak at around 1640 cm -1 and a broad peak at 3340 cm -1 are attributed the stretching vibration mode of OHgroup. S1 Figure S9. XRD patterns of (a) Co 3 O 4 and (b) CoO, and their product after irradiated by different intensity of laser, labeled as -L and -L low corresponding to laser fluence of 700 and 7 mJ/cm -2 , respectively.

Figure S10. Nitrogen sorption isotherms of (a) Co 3 O 4 and (b) CoO, and their product after irradiated by different intensity of laser.
As a comparison experiment, we conducted PLFL on cobalt oxides prepared through direct calcination of cobalt nitrate precursors, which were labelled as C-Co 3 O 4 and C-CoO for the reduced one ( Figure S11). Similar effects on structure could be observed on direct calcined oxides, in comparison to hard-templated cobalt oxides. A small amount of cobalt monoxide was formed on C-Co 3 O 4 ( Figure S12a), which was identified as smaller nanoparticles and contributed a higher surface area ( Figure S12c and S13a). In the case of C-CoO, partial oxidation occcured after laser irradiation ( Figure S12b). According to the refinement result of XRD patterns, 17.2 % of C-CoO was oxidized to Co 3 O 4 spinel ( Figure S14), instead of a much higher oxidation on CoO from hard templating. This can be explained by a shorter distance of oxygen diffusion inside R-Co 3 O 4 nanoparticles (~ 8 nm) during oxidation, compared to C-CoO nanoparticles (20-50 nm). Nevertheless, much smaller particles with 3 times increase of surface area were obtained in the oxidized sample, further supporting that particle fragmentation in water goes along with oxidation process. When employed as OER electrocatalysts, increased activity was obtained on C-CoO after the PLFL process, which is still lower than that of CoO-L ( Figure  S15). Considering oxygen vacancies could also be generated on C-CoO, the superior OER activity of CoO-L results from larger specific surface area and higher ratio of Co 3+ , which is regarded as the active site for cobalt oxides. This result demonstrates the advantage of preparing cobalt oxides from coffee waste templating over solid-solid reaction for the following PLFL process.
To exclude the effect of the laser absorption properties on each sample, ultraviolet-visible spectroscopy was conducted on suspensions of the different cobalt oxide powders ( Figure S16). Although Co 3 O 4 colloids extinct more light at the laser wavelength of 532 nm compared to CoO, the laser absorption is more related to the particle size, since large Co 3 O 4 particles showed even higher laser absorption compared to the small CoO particles. Nonetheless, significant fragmentation was observed in CoO particles despite the lower laser absorption. In summary, it is reasonable to conclude that the oxidation effect is the key factor supporting effective particle fragmentation regardless of particle size. Thermal effects seemed to play only a secondary role by initiating or accelerating the oxidation processes, since comparably low fluence and corresponding low temperatures were required to induce the fragmentation of CoO particles.          In a simplified Randles circuit, R s represents the solution resistance, while C dl element models the double-layer capacitance. The kinetics for Faradaic OER is determined by charge transfer resistance (R ct ).