In-vitro and in-vivo Investigations of Upconversion and NIR Emitting Gd2O3:Er3+,Yb3+ Nanostructures for Biomedical Applications

Eva Hemmer1,*, Hiroyuki Takeshita2, Tomoyoshi Yamano2, Takanori Fujiki3, Yvonne Kohl4, Karin Löw4, Nallusamy Venkatachalam3, Hiroshi Hyodo1,3, Hidehiro Kishimoto1,2 and Kohei Soga1,3

1 Tokyo University of Science, Center for Technologies against Cancer (CTC)

2669 Yamazaki, 278-0022 Chiba, Japan

2 Tokyo University of Science, Division of Immunobiology,

Research Institute for Biological Sciences

2669 Yamazaki, 278-0022 Chiba, Japan

3 Tokyo University of Science, Department of Materials Science and Technology

2641 Yamazaki, 278-8510 Chiba, Japan

4 Department of Cell Biology & Applied Virology,

Fraunhofer Institute for Biomedical Engineering

Ensheimer Straße 48, 66386 St. Ingbert, Germany

* E-mail:

Phone: +81-4-7124-1501-4323

Fax: +81-4-7124-1526


Supporting Information

NIR Emission Detection Limit and NIR Camera Saturation Limit

In order to estimate the detection limits for NIR emission of the used imaging systems, samples of lower as well as higher concentration of Gd2O3:Er3+,Yb3+ particles obtained by CTAB-assisted homogeneous precipitation have been investigated in NIS-Opt and in fluorescence microscope. Therefore, either a diluted particle suspension was dropped on a cover glass (Sample A: approximate concentration of the suspension~2mg/mL, a few drops. Sample B: suspension A was further diluted to an approximate concentration of ~0.25mg/mL, one drop), followed by evaporation of the solvent (ethanol), or a pellet was made by lamination of the dry powder between two polymer foils (sample C). Larger aggregates of dry powders on a glass bottom dish have been used as high concentration sample in case of the fluorescence microscope (sample D). All samples have been excited with 980-nm light and NIR emission images have been recorded as a function of the laser power (respectively the corresponding laser current which is the parameter that is set by the laser control unit). Additional contrast enhancement by use of suitable software (e. g. Corel) improves the image quality in case of low particle concentration. Images obtained by NIS-Opt are shown in Figure 1. Images obtained by fluorescence microscopy are shown in Figure 2.

The NIR emission intensity strongly depends on the particle concentration. In NIS-Opt, an integration time of 200 ms and an aperture of 1.4 were chosen for the imaging of the dispersed samples which correspond to the parameters used for the in-vivo experiments. Yet, due to high NIR emission intensity a lower integration time of 100 ms and an aperture of 16.0 had to be chosen in case of the packed sample in order to avoid damage of the NIR camera. In case of the fluorescence microscope, lower laser power and an integration time of 20ms are sufficient for NIR imaging.

From Figures 1a and 2a it is obvious that NIR emission can be detected from low particle concentrations. As shown in case of Y2O3 nanoparticles, even single particles are detectable in the fluorescence microscope (unpublished data: Akito Hattori, “Size Control of Complexes of Ceramic Nanophosphors and Liposome as Near-Infrared Bioimaging Probes” Master Thesis, Tokyo University of Science).

Figures 1d and 2d show the emission intensity (integrated density) determined with the aid of the software ImageJ (as taken images without contrast improvement have been analyzed). While a saturation limit of the NIR camera could be reached in case of the packed, respectively aggregated sample (Figures 1c and 2c), the emission intensity did not reach a saturation value in case of the dispersed powders, even at the highest laser power. Since the particle concentrations in the in-vivo experiment are comparable to those of the dispersed samples, we assume that the saturation limit of the NIR camera is not reached in our in-vivo studies when the laser power was set to 4.5W. Similar assumptions are made for the fluorescence microscope, which was used to image the histological sections of the organs at a laser power of 0.65mW.


NIR Emission Detection in NIS-Opt

Figure 1. NIS-Opt images of Gd2O3:Er3+,Yb3+ nanoparticles (d=160nm) as a function of laser power. a), b) Particles dispersed on a cover glass (low concentration, integration time: 200ms, aperture: 1.4). c) Packed powders (high concentration, integration time: 100ms, aperture: 16.0). lex=980nm, lem=1.5mm. d) NIR emission intensity (integrated density).

Figure 1 (continued). NIS-Opt images of Gd2O3:Er3+,Yb3+ nanoparticles (d=160nm) as a function of laser power. a), b) Particles dispersed on a cover glass (low concentration, integration time: 200ms, aperture: 1.4). c) Packed powders (high concentration, integration time: 100ms, aperture: 16.0). lex=980nm, lem=1.5mm. d) NIR emission intensity (integrated density).


NIR Emission Detection by Fluorescence Microscopy

Figure 2. a) Fluorescence micrographs of Gd2O3:Er3+,Yb3+ nanoparticles (d=160nm) as a function of laser power. a), b) Particles dispersed on a cover glass (low concentration). c) Powder aggregate (high concentration). lex=980nm, lem=1.5mm, integration time: 20ms. d) NIR emission intensity (integrated density).

Figure 2 (continued). a) Fluorescence micrographs of Gd2O3:Er3+,Yb3+ nanoparticles (d=160nm) as a function of laser power. a), b) Particles dispersed on a cover glass (low concentration). c) Powder aggregate (high concentration). lex=980nm, lem=1.5mm integration time: 20ms. d) NIR emission intensity (integrated density).

2