Nanoscale evidence of erbium clustering in Er-doped silicon-rich silica

Abstract

Photoluminescence spectroscopy and atom probe tomography were used to explore the optical activity and microstructure of Er3+-doped Si-rich SiO2 thin films fabricated by radio-frequency magnetron sputtering. The effect of post-fabrication annealing treatment on the properties of the films was investigated. The evolution of the nanoscale structure upon an annealing treatment was found to control the interrelation between the radiative recombination of the carriers via Si clusters and via 4f shell transitions in Er3+ ions. The most efficient 1.53-μm Er3+ photoluminescence was observed from the films submitted to low-temperature treatment ranging from 600°C to 900°C. An annealing treatment at 1,100°C, used often to form Si nanocrystallites, favors an intense emission in visible spectral range with the maximum peak at about 740 nm. Along with this, a drastic decrease of 1.53-μm Er3+ photoluminescence emission was detected. The atom probe results demonstrated that the clustering of Er3+ ions upon such high-temperature annealing treatment was the main reason. The diffusion parameters of Si and Er3+ ions as well as a chemical composition of different clusters were also obtained. The films annealed at 1,100°C contain pure spherical Si nanocrystallites, ErSi3O6 clusters, and free Er3+ ions embedded in SiO2 host. The mean size and the density of Si nanocrystallites were found to be 1.3± 0.3 nm and (3.1± 0.2)×1018 Si nanocrystallites·cm−3, respectively. The density of ErSi3O6 clusters was estimated to be (2.0± 0.2)×1018 clusters·cm−3, keeping about 30% of the total Er3+ amount. These Er-rich clusters had a mean radius of about 1.5 nm and demonstrated preferable formation in the vicinity of Si nanocrystallites.

Microstructure and optical properties of Pr3+-doped hafnium silicate films

“…Thus, the vibration band at 900 and 1,000 cm −1 can be attributed to Si-O-Pr asymmetric mode. Similar incorporation of rare-earth ions into Si-O bonds and the formation of rare-earth silicate phase was observed earlier for SiO x materials doped with Er 3+ , Nd 3+ , or Pr 3+ and annealed at 1,100°C [17-19]. Thus, based on this comparison, one can conclude about the formation of Pr silicate revealed by FTIR spectra.…”

“…At the initial stage, the annealing process is supposed to decrease the non-radiative recombination rates [24]. Thereafter, the quenching of the Pr 3+ emission that occurred for T A > 1,000°C can be due to the formation of the Pr 3+ silicate or Pr oxide clusters (Figure 2) similar to the case observed in [17,18]. Moreover, it is interesting to note that the position of peak (Pr 3+ : 3 P 0 → 3 H 4 ) redshifts from 487 nm ( T A ≤ 1,000°C) to 492 nm ( T A = 1,100°C) as shown in Figure 4c.…”

On the interplay between Si-Er-O segregation and erbium silicate (Er2Si2O7) formation in Er-doped SiOx thin films

“…For all samples, no signal of Er 3þ ions at 1.54 mm has been detected, evidencing a quenching of the luminescence as usually observed for high doping level of silicon rich silica and high temperature annealing [8,17,21]. Fig.…”

“…Accordingly, an understanding of luminescence requires the control of the structural evolution. However, the study of the nanoscale structure of these systems has only been slightly studied [8,21,22] and consequently the precipitation mechanism of the Er ions in SRSO remains very poorly understood. Hence, our samples were subjected to 1 h heat treatment at 1100 + C, a temperature for which Si-ncs in silica should be formed [12] and thus be able to sensitize efficiently the Er ions [17].…”

Doping porous silicon with erbium: pores filling as a method to limit the Er-clustering effects and increasing its light emission

“…Er-Yb codoping gave better results than the sole Er doping for the PSi matrix 21, 22 , but despite all the efforts made in the rare earth doping of Si structures 23, 24 , no way has been found so far for obtaining light emission with a high enough yield 25, 26 . Er clustering surely plays an important role when trying to increase the availability of Er luminescent centers, and a lot of research effort has been spent on Silicon Rich Oxide (SRO) structures 27–34 , but the clustering is such a hard limiting factor for light emission from these materials that some authors already started to propose to lower Er doping levels 25 to overcome the clustering effect even if it implies reducing the maximum achievable PL emission power.…”