In
the development of oxynitride photocatalysts, thermal ammonolysis
of a metal oxide precursor has often been conducted by varying the
reaction conditions (e.g., temperatures, reaction times, and ammonia
gas flow rates) to obtain high-quality oxynitride particles that efficiently
function as photocatalysts. However, this approach may suffer from
undesirable changes in the physicochemical properties of the resulting
oxynitride, leading to the lowering of the photocatalytic activity.
Here, we show that it is possible to control the photocatalytic activity
of Ruddlesden–Popper metastable layered oxynitride K2LaTa2O6N, obtained from the Dion–Jacobson
phase KLaTa2O7 through a topochemical ammonolysis
reaction, by controlling the quality of the KLaTa2O7 template. During the ammonolysis of KLaTa2O7, in the presence of K2CO3, to K2LaTa2O6N, the structural properties
(e.g., degree of crystallinity and particle size) of the oxide precursor
were replicated in the resulting oxynitride. Namely, the use of KLaTa2O7, possessing a higher degree of crystallinity,
led to larger K2LaTa2O6N particles
being formed. By increasing the crystallinity of KLaTa2O7, the photocatalytic activity of the resulting K2LaTa2O6N for H2 evolution
was improved for reaction in aqueous NaI solution under visible light
irradiation. This improvement in performance was due to the longer
lifetime of the photogenerated mobile electrons in high-crystallinity
K2LaTa2O6N compared with that in
the low-crystallinity analogue, as confirmed by femtosecond transient
absorption spectroscopy. However, the photocatalytic activity of K2LaTa2O6N derived from well-grown larger
KLaTa2O7 particles was an order of magnitude
lower than that of the best-performing material. Physicochemical measurements
revealed that the large K2LaTa2O6N particles contained a relatively high density of anionic defects
on the surface, which shortened the lifetime of the photogenerated
charge carriers, leading to lower photocatalytic activity.