Impact of ultrasonic dispersion on the photocatalytic activity of titania aggregates

The effectiveness of photocatalytic materials increases with the specific surface area, thus nanoscale photocatalyst particles are preferred. However, such nanomaterials are frequently found in an aggregated state, which may reduce the photocatalytic activity due to internal obscuration and the extended diffusion path of the molecules to be treated. This paper investigates the effect of aggregate size on the photocatalytic activity of pyrogenic titania (Aeroxide® P25, Evonik), which is widely used in fundamental photocatalysis research. Well-defined and reproducible aggregate sizes were achieved by ultrasonic dispersion. The photocatalytic activity was examined by the color removal of methylene blue (MB) with a laboratory-scale setup based on a plug flow reactor (PFR) and planar UV illumination. The process parameters such as flow regime, optical path length and UV intensity are well-defined and can be varied. Our results firstly show that a complete dispersion of the P25 aggregates is not practical. Secondly, the photocatalytic activity is not further increased beyond a certain degree of dispersion, which probably corresponds to a critical size for which UV irradiation can penetrate the aggregate without significant obscuration.

S3 -In the PFR, the process is assumed as a steady-state isothermal reaction.
-Since the reaction is slow and dominates over the mass transfer, the kinetic of PFR is considered by means of a pseudo-homogenous model.
-Reaction occurs only in the PFR where photocatalyst is activated by absorbing photon energy of UV-irradiation and it is a first-order reaction -There is no reaction in the mixing tank.
-The mixing in the tank is assumedly ideal.

Modeling
Consider a volume element of species A in the PFR, the material balance follows Since the reaction is first-order, 1 ϑ = − , and 1 r kC = .
The integral gives the solution as Material balance for species A in the mixing tank is described as Figure S1: Solution of the characteristic equation.
The boundary condition is determined at Accordingly, the conversion of species A follows and the reaction rate constant k is investigated as

Calibration curve of methylene blue measured by UV-vis spectroscopy
Methylene blue (MB) (Merck KGaA) 3.126 mM was used as the stock solution for the calibration. The eight concentrations ranging from 0.537−12.344 µM were prepared.
The absorbance of solutions was scanned in the wavelength of 200−800 nm through 10 mm optical path length by a Varian Cary 100 Bio spectrometer. The maximum absorbance at wavelength λ = 664 nm was picked out from the whole UV-Vis spectra.
All measurements were repeated three times. Analysis of the values along with MB concentrations were shown in Figure S2, where standard deviations of each point are too small to be observed, lower and upper CI respectively show the 95% confidence intervals of the linear fit. Figure S2: Calibration of methylene blue.

Ultrasonic dispersion of P25 aggregates
The 1 g/L P25 suspensions were disintegrated by ultrasonication. Aggregate sizes were examined as a function of energy density E V (Table S1-Table S6 where parameters were determined by correlating measurements along with time t by two instruments (HPPS-ET and Nano S90).

Stability of dispersed suspensions
The stability of the P25 titania suspensions (at room temperature) were examined by monitoring the size parameters over a time period of three days. Their aggregate sizes measured by PCS Malvern Nano S90 including cum x , PDI , 50,int x , 50,0 x , and 50,3 x (Table   S7) show that a stable suspension can be achieved by ultrasonication, while the reaggregation of the colloids is induced by a conventional stirring.

Color removal of MB in P25 suspensions
The photocatalytic properties of P25 were examined by the discoloration of MB.
Photocatalyst aggregate sizes were measured by PCS Malvern Nano S90 including cum x , PDI , 50,int x , 50,0 x , and 50,3 x . The 90% quantile of intensity-weighted cumulative distribution 90,int x calculated as Eq. (7) is considered as agglomerate size of photocatalyst. The apparent reaction rate constant K in the whole setup and the intrinsic reaction rate constant k in the reactor are calculated as Eq. (3) and Eq.(4). Data are given in Table S8 and Figure S3.