In operando imaging of self-catalyzed formaldehyde burst in methanol oxidation reactions under open circuit conditions

A wave-like HCHO burst is in situ observed during electro-oxidation of methanol on Pt under open circuit conditions by SPR imaging.


Preparation and characterization of PtNPs
The platinum nanoparticles (PtNPs) are synthesized through a rapid and one-step method according to previous literature. 1 Briefly, the PtNPs are prepared by mixing H 2 PtCl 6 (3 mL, 18.9 mM), ascorbic acid (0.2 mL, 1 M) and NaOH (15 μL, 5 M) together in a beaker which contains 18 mL H 2 O. After heating the mixture in a water-bath at 60 o C for 10 min without any disturbance, the brown suspension containing monodisperse PtNPs are obtained. Such surfactant-free PtNPs possess definitely "clean" surfaces and can be directly used for subsequent experiments after simple washing with water. The morphology and structure of the as-synthesized sample are characterized by TEM (JEM-2100) and SEM (Shimadzu, S4800). Fig. S1 shows that the PtNPs are almost monodisperse with a mean particle size of 500 nm, which is further confirmed by the dynamic light scattering (DLS, Malvern Nano-ZS90 Zetasizer) data. Additionally, the high-resolution TEM image shows that the PtNPs are in the polycrystalline state, of which the d spacing is 0.22 nm and 0.19 nm, which can be assigned to be [111] and [200] planes of Pt, respectively.

SPR imaging (SPRi) and image processing
The prism-based SPRi setup is comprised of an LED light source, collimating lens, prism, polarizer and a high-speed charge-coupled device (CCD) camera (Pike, Allied Vision Technologies). Specifically, the p-polarized red light from the LED (wavelength 670 nm) is collimated and directed to the sensor chip via a triangular prism at an incident angle near SPR resonance angle. The SPR sensor chip is a BK7 glass cover slide coated with 2 nm chromium followed by 47 nm gold, which is also used as the working electrode.
Before use, each chip is washed with water and ethanol, followed by briefly annealing with a hydrogen flame to remove surface contamination. 0.2 μL of the PtNPs solution is dropped onto the chip surface via the pipette, followed by dried in a vacuum. Then this Au chip, used as the SPR sensing surface, is put upon the prism using indexmatching fluid between them. A PDMS chamber (Flexiperm micro12, Sarstedt) is placed on top of the gold surface for housing the reaction solution and electrodes. As to electrochemical system, both the counter electrode (CE) and reference electrode (RE) are inserted to the electrochemical cell from the top opening. SPRi images are recorded at a speed of 7.6 frames per second (fps). Then the intensity is determined by calculating the average intensity within a fixed small rectangle region-of-interest (ROI) selected close to the Pt-disk.

Spectrophotometric assay of HCHO
Determination of formaldehyde is accomplished by a spectrophotometric method with a methylene blue-bromate (MB-KBrO 3 ) system. All solution is kept in a 30 ℃ thermostat water batch for 20 min before initiation of the reaction. To a series of 5 mL tubes, 40 μL of 0.05 mM MB solution, 0.6 mL of 0.5 M H 2 SO 4 (or electrolyte after electro-oxidation), and different amounts of HCHO are added. Then 100 μL of 10 mM bromate solution is added and the solution is diluted to 1.8 mL with deionized water.
After shaking, the mixed solution is transferred into quartz cell for subsequent measurement by a spectrophotometer (Cary 100, Agilent Technologies). The variation in absorbance of MB (λ max = 664 nm) versus time is continuously monitored after the addition of bromate solution. The blank reaction is performed according to the same procedure without addition of HCHO and the absorbance is labeled as A 0 . The standard curve which the difference of absorbance between blank and sample in 664 nm (A 0 -A) of MB versus the concentration of HCHO is also constructed.

Electrocatalytic activity of the Pt-disk chip for methanol electro-oxidation
A CHI 760E electrochemical workstation (Shanghai Chenhua Instrumental CO., Ltd., China) is employed for all the electrochemical experiments with a Pt wire and Ag/AgCl served as the counter electrode (CE) and the reference electrode (RE), respectively. The SPR sensor chip is also used as the working electrode (WE).
Pre-synthesized colloids containing 500 nm spherical Pt nanoparticles are deposited on gold-coated coverslip to form a porous and disk-like Pt-catalyst (Pt-disk) with a diameter of ~1000 microns. Gold film acts as both the working electrode and the optical interface to excite surface plasmon polaritons. Electrochemical characterization in the presence of methanol displays a pair of well-documented oxidative current peaks in the cyclic voltammogram ( Typically, the peak corresponding to the formation of platinum oxide appears after 0.6 V, and another peak that corresponding to the reduction appears at 0.48 V. However, when the electrolyte contains the methanol, the significant redox peaks of methanol oxidation appears at 0.56 V in the positive-going scan and 0.45 V in the negative-going scan, respectively, demonstrating the reliability of such Pt nanoparticles covered Au chip as working electrode in the electrocatalysis of methanol (Fig. S2, red line). The asprepared Pt-disk chip exhibits the equivalent catalytic performance due to the relatively lower anodic overpotential.

3-Dimensional semi-infinite diffusion model
According to the diffusion model, we assume an immediate (t = 0) and point (x = 0, y = 0, z = 0) release of M molecules. These molecules diffuse at a semi-infinite space (-∞ < x < +∞, -∞ < y < +∞, 0 ≤ z < +∞). Diffusion coefficient (D) is homogenous at any space location. Then, the basic equation is: where H is the height of source point, and H = 0 here. Normally SPR measures the concentration distribution in the basal plane of this semi-infinite diffusion space. Penetration depth of surface plasmon wave (200 nm) is significantly smaller than the diffusion distance (r), which is often in the scale of 100 microns. So, we actually measure the concentration distribution in the basal plane (z = 0). If we solely focus on the basal plane and let distance , then At each location, the initial concentration (C(r,0) = 0) is zero. Subsequently, it will increase due to the species diffused from the source. Eventually the concentration would decrease and approach zero again. The time required to reach the maximal concentration is a function of distance (r) to the center of the Pt-disk. It takes shorter time to reach the maximal concentration for a point with shorter distance. The time is defined by the first order derivative of dC/dt. . Together with the experiment data, the results follow the relationship = 6 × 1 2 that , in which the slope (k = ) of this linear fitting equation is = 6 × 1 2 6D 1.7×10 -4 . Thus, the diffusion coefficient D is 4.8×10 -9 m 2 /s.

Reaction mechanism of MB-bromate system for formaldehyde detection
Methylene blue (MB) is a dye that can be oxidized with strong oxidizing agents at slow reaction. Formaldehyde can accelerate the reacting rate at ultra-trace level, and the catalytic effect increased with increasing concentration of the formaldehyde. [2][3][4][5] The mechanism of the reaction between MB and potassium bromate may attributed to the equations as below:  4)]. This reaction is monitored spectrophotometrically by measuring the decrease in absorbance of MB at about 664 nm, which can be used to measure the concentration of formaldehyde.
Thus, we take this method to further identify the formaldehyde produced in methanol oxidation, and the absorbance spectra of different solutions by UV-vis spectrometry (Cary 100, Agilent Technologies) are displayed in Fig. S4 for checking its reliability. The reaction rate of the MB with KBrO 3 in sulfuric acid is very slow and the absorbance at about 664 nm is still strong (Fig. S4-a, curve 1). However, the rate of MB oxidized by KBrO 3 increases quickly in the presence of HCHO (curve 4) and the peak absorbance decreases in large range comparing with the curve 1. We note that no pronounced catalytic effect can be find in the CH 3 OH or HCOOH system. The kinetic curves for the absorbance variation as a function of time can give a better illustration and it is demonstrated that MB could be quickly oxidized by bromate ion in the presence of both determinate HCHO solution and real electro-oxidation sample ( Fig. S4-b). Thus, it is reliable to use this selective method to detect the real sample of methanol oxidation without having to consider the interference from the reactant methanol and intermediate formic acid.
After electro-oxidation of the H 2 SO 4 -CH 3 OH solution at different constant potentials for 30 s on the Pt-dot chip, the electrolytes are collected and then sampled with MB-KBrO 3 solution for subsequent online analysis by UV-vis spectroscopy. We also collect the different absorbencies of freshly prepared standard solutions in the range of 0.3-100 μM. The linear calibration plot in Fig. S4-c demonstrates the assay sensitivity and precision. The limit of detection as near 200 nM. As a result, the produced HCHO quantity under different applied pre-oxidation potentials is determined from the absorbance change in comparison to the calibration curve.

The refractive index of formaldehyde and methanol.
We have the SPR intensity change when adding different amount of HCHO and CH 3 OH into the 0.5 M H 2 SO 4 +1.0 M CH 3 OH electrolyte (EL-1 solution), respectively.
The corresponding refractive index (RI) of HCHO and CH 3 OH solutions are also investigated at the same time. The RI of single pure formaldehyde solution (37%, v/v) is measured as 1.377 at the temperature of 15 ℃ by WAY-2S ABBE Digital Refractometer (Shanghai INESA Physico-Optical Instrument CO., Ltd, China). As to methanol solution (99.9%, v/v), its RI value is 1.331. When we add CH 3 OH and HCHO into the 0.5 M H 2 SO 4 +1.0 M CH 3 OH electrolyte (EL-1) solution, respectively, the RI of mixed solution gradual increases with the increasing concentration (Fig. S6). And addition of HCHO has higher rising rate, which is consistent with the conclusion of SPR intensity variation in Fig. 2d. Thus, SPRi can easily catch the emergence of HCHO, as well as the concentration change of HCHO, due to the different RI.

The electrochemical current and SPR intensity in the CA measurement.
The electrochemical current curves in the stage 4 of CA measurement are also recorded and displayed in Fig. S8. We find that the current value in H 2 SO 4 -CH 3 OH solution is always larger than that in corresponding separated H 2 SO 4 solution. And the current values of Pt-dot in H 2 SO 4 -CH 3 OH subtracted by that in H 2 SO 4 are all positive values, which can be viewed as oxidation currents, indicating the generation of formation in this stage 4 actually derives from the oxidation of methanol indeed. If we integrate the differential current over time, then we will get the total amount of electric charge inside the stage 4.
As to the SPR intensity, the low-potential of 0.4 V is an exception and discussed in Fig. 3b. For the low-potential from 0 V to 0.3V, the amount of generated formaldehyde has a minimum at about 0 V, which is close to then OCP of system (-0.05 V). The more far away the OCP, the more electrons transfer at the chip/electrolyte interface and contribute to oxidize the methanol to formaldehyde. In other words, under the premise of adsorbing same amount of Pt oxides when the applied low-potential is very close to OCP, it only needs less driving force (i.e. electrons) to return to the system steady state and the production of formaldehyde decreases accordingly. However, it's important to point out that too negative cathode potential (e.g. -0.2 V) was unfavorable to the oxidation of methanol molecule, and no obvious SPR intensity peak appeared.     (1)

Time (s)
in H 2 SO 4 in H 2 SO 4 +CH 3 OH

Fig. S8
The corresponding current curves of the Pt-disk chip in chronoamperometry measurement when the applied potential is stepped from 0.5 V to and hold at lower potential of 0 V. The red line is the current from the H 2 SO 4 + CH 3 OH solution, while the black line is from the H 2 SO 4 solution.