submitted at DPC'03, Christchurch, NZ for J. Lumin.

Long-lifetime luminescence of lanthanide-doped gadolinium oxide nanoparticles for immunoassays

W.O. Gordon, J. A. Carter, and B. M. Tissue*
Department of Chemistry, Virginia Polytechnic Institute and State University
Blacksburg, VA 24061-0212, USA

Abstract

This paper reports the luminescence spectra and lifetimes of lanthanide-doped Gd2O3 nanoparticles to identify long-lifetime luminescent reporters for immunoassays. The particles were prepared by gas-phase condensation and then annealed to obtain crystalline single-phase particles. Tb3+, Dy3+, and Eu3+ provide strong emission lines at 544, 573, and 611 nm, respectively. Of these three dopants, Dy3+ was the most sensitive to concentration quenching, but retained a lifetime of 120 µs at a concentration of ~1 %.

Introduction

The small size of luminescent inorganic nanoparticles allows them to replace fluorescent molecules or complexes in analytical applications [1]. One example is as a luminescent tag or reporter for affinity or immunoassays in environmental, food quality, drug, and biomedical testing. In these assays a luminescent reporter is attached to an antibody that binds to a specific analyte or antigen. Detection or quantitation of the analyte is then based on luminescence intensity (or quenching), energy transfer, polarization anisotropy, etc. The long lifetime of lanthanide ions provides a significant advantage in discriminating against background autofluorescence [2], and using lanthanide-doped nanoparticles permits the use of NIR-upconversion excitation to avoid autofluorescence completely [3].

In addition to the long lifetime, other considerations for applying luminescent nanoparticles in multi-spectral immunoassays include: (1) finding multiple reporters with distinct emission wavelengths to attach to different antibodies, (2) having a common excitation wavelength for the different reporters, and (3) keeping the particles sufficiently small or soluble to not alter the activity or solubility of the antibodies. Despite the advantages of luminescent nanoparticles, there are several challenges to their use in analytical applications: (1) inherent size effects can cause inhomogeneous spectral broadening due to disorder, metastable structures, or heterogeneous dopant distribution, and (2) the luminescence dynamics can be sensitive to changes in the electron-phonon interaction, phonon spectrum, surface quenching, metastable population, and spontaneous transition rate. This paper presents our first survey experiments to find lanthanide-doped nanoparticles that will be suitable as immunoassay reporters. We use Gd2O3 as the host to take advantage of the ultraviolet excitation with efficient energy transfer to the dopant: Pr3+, Dy3+, Eu3+, Tb3+, and Tm3+.

Experimental

Mixtures of Gd2O3 and Ln2O3 (Ln = Eu, Dy, Tm) were pressed into a pellet and sintered overnight at 700oC. Tb and Pr were introduced into Gd2O3 as the oxalate (Ln2(C2O4)3) to maintain the 3+ oxidation state. The oxalate is synthesized by mixing 50 mL of 0.2 M aqueous oxalic acid at 80oC with 100 mL of 0.025 M Ln(NO3)3 also at 80oC. The resulting Ln2(C2O4)3 precipitate is rinsed, filtered and dried before pressing with Gd2O3 and sintering at 500oC for 5 hours. During vaporization of these samples, the presence of residual oxalate provides a reducing atmosphere to help maintain the Ln ions in the 3+ oxidation state.

The nanoparticles were prepared by gas-phase condensation using a cw-CO2 laser to vaporize from the sintered ceramic pellet [4]. The target rotated on a platform in a vacuum chamber filled with 1 Torr N2 and nanoparticles collected on a stainless steel collector (see Figure 1). The as-prepared nanoparticles had a fluffy morphology and were white for Eu, Dy, and Tm, but were tinted brown for Tb and Pr. This discoloration could be indicative of oxidized dopant ions or residual carbon. The nanoparticles were placed in platinum crucibles and annealed at 700o C for 1 hr to improve the crystallinity and to obtain single-phase material. The spectra are labeled with the dopant concentration of the target, but the concentration in the nanoparticles can be different depending on the preferential vaporization of the dopant or host oxides. Transmission electron micrographs were obtained using a Phillips EM 420 transmission electron microscope (TEM) operated at 100kV to determine particle diameter.


Figure 1. Schematic of the gas-phase condensation nanoparticle synthesis chamber. The laser power is typically 30 W and the target-to-collector distance is 3-5 cm. Two additional end plate collectors are not shown.

For optical spectroscopy, powder was packed into a 3-mm diameter indentation in a copper block mounted inside a vacuum chamber. All spectra were taken with the sample under vacuum (~ 3x10-2 Torr) at room temperature using pulsed excitation at 266 nm (Continuum Nd3+:YAG laser) and a 1-m monoChromator (Spex 1000M) with a bandpass of 0.3 nm or less. The detector was a Hamamatsu R-636 PMT and spectra and decay transients were recorded with a gated photon counter (Stanford SR400) and computer data acquisition. Some transients were also recorded by signal averaging with a 500-MHz digital oscilloscope (Tektronix TDS 520A).

Results and Discussion

Annealing - The spectra of the as-prepared (unannealed) nanoparticles showed broadened peaks characteristic of multiple disordered phases with less efficient emission at the desired wavelength. Figure 2 shows spectra of Eu3+:Gd2O3 as-prepared and annealed at 700 and 800 oC. The as-prepared sample shows broad peaks that indicate the presence of monoclinic and disordered phases [5]. Annealing converts the nanoparticles to the cubic phase and greatly improves crystallinity. TEM showed that as-prepared nanoparticles ranged in size from 5-10 nm. Annealing in air at 700 oC for 1 hr increased particle size to 10-15 nm and annealing for 4 hours at 800 oC increased the particle size to 20-25 nm. Figure 2 shows that the 1 hr anneal at 700 oC produces the desired improvement in the luminescence spectra while minimizing grain growth and a longer anneal at higher temperature has little additional effect on the luminescence.


Figure 2. Comparison of luminescence spectra of 0.1 % Eu3+:Gd2O3 as-prepared and annealed in air. The spectra were recorded with a 1-ms delay and 1-ms gate. The intensity scale is arbitrary.

Eu3+:Gd2O3 - Annealed Eu3+:Gd2O3 nanoparticles show strong 5D0 --> 7F2 emission at 611 nm as seen in Figures 2 and 3. The lifetimes of unannealed 0.1, 1, and 5 mol-% Eu-doped Gd2O3 were 200, 290, and 170 µs, respectively. These lifetimes are much shorter than expected, but probably reflect the presence of Eu2O3 in the samples. The lifetime of these samples increased to 1650, 900, and 670 µs, respectively, after annealing at 800 oC for 4 hours. The lifetimes are longer than in bulk material due to the effect of the surroundings [6], but nonradiative effects become the determining factor at the highest concentration.


Figure 3. Luminescence spectra of lanthanide-doped Gd2O3 nanoparticles. The delay and gate are the same as in Figure 2 for Eu3+ and 5 µs and 300 µs, respectively, for the Tb3+ and Dy3+ spectra. The intensity scale is arbitrary.

Tb3+:Gd2O3 - Figure 3 shows the luminescence of annealed nanoparticles vaporized from 3 % Tb3+:Gd2O3. The strongest line at 544 nm corresponds to the 5D4 --> 7F5 transition. Although the terbium sample has significant secondary emission lines relative to the 544-nm line, these other lines do not overlap the strong lines of Dy and Eu. Tb3+:Gd2O3 nanoparticles annealed at 700 oC for one hour with concentrations of 0.1 %, 3 %, and 5 % Tb had lifetimes of 920, 330, and 250 µs, respectively. Of these three samples, the 3 % doping level provides the best compromise between long lifetime and highest luminescence intensity.

Dy3+:Gd2O3 - The luminescence spectrum of Dy3+:Gd2O3 annealed at 700 oC for one hour showed its strongest emission at 573 nm (see Figure 3), corresponding to the 4F9/2 --> 6H13/2 transition. An annealed 1 % sample had a lifetime of 120 µs, but samples of 2 % and 5 % Dy were strongly quenched with lifetimes of 45 and 20 µs, respectively. The dysprosium samples have somewhat lower luminescence intensity than the corresponding terbium or europium samples, but probably have sufficient intensity for immunoassay applications.

Other Dopants - We also investigated Pr3+ and Tm3+ doped Gd2O3. The spectra of Pr3+-doped Gd2O3 nanoparticles showed a weak emission at 510 nm and a broad emission centered at 620 nm. This red emission overlaps with the Eu emission line, and the green emission is too weak to be useful, unlike micron-sized particles [7]. In addition, the lifetimes of the green and red emission lines were short, ~ 4 µs and 45 µs, respectively, in the 0.1 % sample. Blue luminescence was observed in one 5 % Tm3+-doped Gd2O3 sample, but not in other samples (5 %, 1 %, 0.1 %). The thulium luminescence spectrum showed several weak peaks of similar intensity from 460-520 nm with short lifetimes (20-30 µs). We believe that the thulium does not vaporize as well as the gadolinium or other lanthanide oxides using our laser-heated condensation method.

Conclusion

These survey experiments of 5 different lanthanide ions doped into Gd2O3 nanoparticles demonstrate host and dopant combinations with a common excitation wavelength that can be used for multicolor immunoassays. Our preliminary results show that suitable dopants with long lifetimes and distinct emission wavelengths include 3 % Tb3+, 1 % Dy3+, and 1-5 % Eu3+. The gas-phase condensed nanoparticles require a 1 hr anneal at 700 oC to produce crystalline single-phase material, which might require aggressive treatment to disperse agglomerates for coupling to an antibody. Other issues to investigate in the future are suitable coupling chemistry, surface passivation, and the incorporation of co-dopants to use NIR-upconversion excitation.

Acknowledgments
We thank S. McCartney for assistance with TEM.

References
[1] T. Soukka, H. Harma, J. Paukkunen, T. Lovgren, Anal. Chem. 73 (2001) 2254.	
[2] I. Hemmila, V.M. Mukkala, Crit. Rev. Clin. Lab. Sci. 38 (2001) 441.
[3] R.S. Niedbala, H. Feindt, K. Kardos, T. Vail, J. Burton, B. Bielska, S. Li, D. Milunic, P. Bourdelle, R. Vallejo, Anal. BioChem. 293 (2001) 22.
[4] H. Eilers, B.M. Tissue, Mater. Lett. 24 (1995) 261.
[5] B.M. Tissue, H.B. Yuan, J. Solid State Chem. 171 (2003) 12.
[6] R.S. Meltzer, S.P. Feofilov, B.M. Tissue, H.B. Yuan, Phys. Rev. B 60 (1999) R14012.
[7] M. Okumura, M. Tamatani, A.K. Albessard, N. Matsuda, Jpn. J. Appl. Phys. 36 (1997) 6411.