2. Experimental

The experiments are summarized into three major parts. First, impurity contents that may affect the EL properties of currently the best SrS:Ce based thin films prepared by Atomic Layer Epitaxy (ALE) and reactive evaporation were investigated by combined ion beam analysis techniques. Second, ALE deposited SrS, SrS:Ce and SrS:Cu thin films were modified by implantation of selected ions and their luminescence properties were studied. Related to this, to understand the role of chlorine in the luminescence of SrS:Ce, study was made of the luminescence of cerium in powder alkaline earth chloride hosts. Third, PL and EL of the recently developed SrS:Cu and SrS:Ag,Cu materials were examined.

2.1. Sample preparation

For the experiments, SrS based thin films of different structure were prepared by several deposition techniques. Typically, SrS phosphor thin films having a thickness of about 500 nm were deposited on glass substrates coated with ITO transparent electrodes underneath an Al2O3 insulating layer. Occasionally, silicon wafers were used as substrates. For ion beam analyses of impurity contents in the phosphor layer, the structures were covered in situ or ex situ with 10, 35, or 100-200 nm of a protecting Al2O3 layer. Some samples also had an in situ deposited ZnS protective layer (about 100 nm). For ion implantation studies, 500-nm-thick ALE SrS films with Ce doped in the top 100-nm layer were used, as well as fully Ce doped films. An in situ 10-nm Al2O3 layer was deposited on top of the SrS layer to protect the phosphor from moisture and contamination. To allow EL measurements of the implanted samples, a further 200 nm Al2O3 top insulation layer was deposited after a heat treatment of the samples. In the case of the SrS:Cu and SrS:Ag,Cu samples used for ion implantation and/or luminescence studies, the as deposited phosphor films were first subjected to heat treatment without a protecting layer, and the top Al2O3 insulating layer was deposited afterwards to complete the EL structures.

The following text briefly describes the phosphor deposition techniques employed in this study.

ALE, also known as ALD (atomic layer deposition) and AL-CVD (chemical vapor deposition), and under some other names, is one of the major deposition techniques for producing high quality commercial TFEL displays. The technique is based on sequential surface reactions of different chemical precursors in a self-limiting manner, and the thin film is grown in layer-by-layer fashion with a precise thickness and high conformality.(8) Crystalline SrS:Ce thin films are deposited by using Sr(thd)2, Ce(thd)4 or Ce(cp)3, and H2S precursors at a substrate temperature of about 380°C.(9), (10) (thd=2,2,6,6-tetramethyl-3,5-heptanedione, cp=cyclo-pentadenyl) Except for the electrodes, the whole EL structure is prepared in vacuum with a single pump-down. After a necessary thermal treatment at about 515 ° C, excellent EL performances with L40 of >90 cd/m2 at 60 Hz driving condition can be routinely obtained. The ALE technique is also able to produce SrS:Cu thin films with both blue and green EL with Cu(thd)2 as Cu precursor. Suitable Ag precursors for ALE are unavailable, however, and codoping of Ag into SrS:Ce and SrS:Cu by ALE has yet to be realized.

In reactive evaporation, SrS:Ce,Cl thin films are deposited in ultra-high vacuum at substrate temperature above 600 °C with elemental Sr, S, and powder CeCl3used as source materials. High luminance and efficiency can be achieved by the addition of elemental Mn.(11)The advantages of reactive evaporation technique are the flexibility to choose the reactants and to control the stoichiometry of the thin film. EL performances have been further improved by the addition of ZnS as the result of filling the S vacancies.(12) The as deposited films are highly crystallized and no further heat treatment is necessary. However, preparation of an EL device requires deposition of a top insulating layer in a separate step. Recently, an even further enhancement of EL performance has been made by codoping with Ag. Reactive evaporated SrS:Ce,Ag,Mn,Cl reached L50 of 170 cd/m2at 60 Hz driving condition.(13) This is the highest luminance ever achieved for SrS:Ce based materials.

Sputtering is a frequently used thin film deposition technique. The use of single target source material and the high energy of sputtered atoms during the deposition process make it a feasible method for producing films with complex composition while maintaining good stoichiometry. The new SrS:Cu and SrS:Ag,Cu blue phosphors studied in this work were deposited by rf magnetron sputtering at substrate temperature 150-350°C.5,7 Owing to the intrinsic characteristics of the sputtering process, the crystallinity of the as deposited films is poor, but it can be greatly improved by post deposition RTA at about 800 °C. The top insulating layer was deposited after the heat treatment.

Several SrS:Cu samples were also prepared by molecular beam epitaxy (MBE). The technique is well known to produce high quality thin films, but it has not been employed in TFEL device production owing to the high cost of the equipment. Similar to ALE and sputtered SrS:Cu films, MBE SrS:Cu films require post deposition annealing at 800 °C.(14)

2.2. Ion beam analysis techniques

Ion beam analyses are based on interactions between a bulk material and energetic ion beams produced by an accelerator. With a suitable instrument set-up, information on element content and a depth profile of the target sample can be obtained through methods of detection that are specific to individual elements. These methods include back scattering (BS), elastic-recoil-detection analysis (ERDA), particle-induced x- or g-ray emission (PIXE, PIGE), nuclear reaction analysis (NRA), and nuclear resonance broadening (NRB). Most of the methods used in this study employed a 2.4 MV Van de Graaff accelerator which is able to generate 1H+, 2H+, and 4He+ ion beams. For the Time-of-Flight (TOF)-ERDA, ion beams of 197Au7+ or 127I6+ were obtained from a 5 MV EGP-10-II tandem accelerator.

Table I. Comparison of ion beam techniques used for SrS:Ce thin film impurity analysis.

Table 1

The advantages of ion beam methods in thin film analysis are as follows: capable of detecting most of the elements down to hydrogen, quantitative, capable of depth profiling in a range of a few nanometers to a few micrometers, and non-destructive nature. Thin film characterization is by no means an easy task, however, because each ion beam method has its own limitations and the composition and structure of thin films are often complicated. The diversity and complexity of trace impurities in thin films impose an even greater challenge for quantitative analysis. Fortunately, combined ion beam techniques are able to provide at least qualitative information. With effort, the impurity contents in a thin film can be detected with reasonable precision. Table I lists several ion beam techniques used for the impurity analysis of SrS:Ce based thin films, together with their capabilities in element range, detection limit and depth profiling, and their application in this work.

2.3. Ion beam modification

Ion implantation is an important technique widely used in the semiconductor industry. The simple process and the capability to implant almost any element in a controlled quantity also make it a very attractive tool for TFEL material studies. With the aid of ion beam analysis and PL and EL studies, the influences of different impurities on the luminescence properties and the mobility and concentration of dopants and codopants can be rapidly examined. In this case, the possibility to implant one type of ion at a time is advantageous as compared with codoping using chemical compounds, where the counter-ions are also codoped and it is difficult to know which ions are responsible for the observed effects. Furthermore, EL devices can be made from ion beam modified thin films.(15)

In this work, ion implantation of various elements into ALE grown SrS, SrS:Ce, and SrS:Cu thin films was carried out using an isotope separator with a maximum energy of 120 kV. The energies for each element were adjusted to obtain an implantation depth of about 100 nm. The wide range of the implantation doses correspond to average concentrations of 0.003-1.0 at.%. The dopant concentrations in normal EL devices usually are in the range of 0.1-0.5 at.%. To avoid complication, problems such as implantation profile, implantation-caused non-stoichiometry, and the effect of oxidation states of non-metal elements were not taken into consideration. Since implantation causes lattice damage to the phosphor, annealings at 500-800 °C in N2 atmosphere for various periods (typically 2 hours) were performed for all implanted films.

2.4. PL and EL measurements

Photoluminescence measurement is one of the major methods employed for study of the luminescence properties of SrS based blue phosphors. Not only does the PL emission closely resembles the EL, but more importantly, PL excitation, decay and time resolved spectra studies are very helpful for understanding the luminescence mechanism of the phosphor. Many luminescence properties can be further explored with use of the low temperature facility.

Ultimately, luminance, efficiency, and CIE color coordinates of the films must be characterized through EL measurements. In this work, EL spectra and L-V curves were studied using an ac power supply operating at 60 Hz driving conditions.

Figure 5

Fig. 3. PL and EL CIE color coordinates of a set of ALE SrS:Ce TFEL devices obtained with different excitation methods.

Quantitative comparison of the PL and EL properties of phosphor thin films is not always straightforward, however. A study on CIE coordinates of a set of SrS:Ce samples showed that the results do not correlate with each other when they are excited at Ce center (430 nm, PL), via host (270 nm, PL), or by EL excitation. (Fig. 3) Another frequently encountered problem is the change of the emission spectrum due to optical interference within the thin film stack that consists of several layers having different refractive indices and a reflecting substrate or electrodes.(16) The interference is particularly strong when the substrate is Si or, in the case of EL measurement, the back electrodes are metallic. PL and EL properties from different samples may be compared, therefore, only if the thin film stacks are of identical structure and thickness, and have the same measurement geometry, even though the interference effects cannot be avoided by this way. To eliminate interference experimentally, one has to be able to integrate the light from the whole emission sphere. Yet another problem is the optical scattering originating from an opaque film, which may increase the measured PL and EL intensity. Also other factors should be taken into consideration, such as the phosphor film thickness and the amount of transferred charge in EL characterization (DQ, defined as total charge per unit area being transported across the phosphor layer between the two polarities). To compare the luminance and efficiency of ALE SrS:Ce and reactively evaporated SrS:Ce,Mn,Cl, the transferred charge is specified to 1 mC/cm2.

8. T. Suntola, in Handbook of Crystal Growth 3, (ed. D.T.J. Hurle), Elsevier, 1994, pp. 605.

9. E. Soininen, M. Leppänen, and A. Pakkala, 13th Inter. Display Research Conf., Eurodisplay'93, Strasbourg, 1993, pp. 233.

10. G. Härkönen, M. Lahonen, E. Soininen, R. Törnqvist, K. Vasama, K. Kukli, L. Niinistö, and E. Nykänen, 4th Inter. Conf. Sci. Tech. Display Phosphors, Extended Abstracts, Bend, OR, 1998, pp. 223.

11. B. Hüttl, K.O. Velthaus, U. Troppenz, R. Herrmann, and R.H. Mauch, J. Cryst. Growth 159 (1996) 943.

12. R.H. Mauch, K.O. Velthaus, B. Hüttl, U. Troppenz, and R. Herrmann, SID 1995 Digest 26, pp. 720.

13. T.A. Oberacker, U. Troppenz, B. Hüttl, T. Gaeryner, R. Herrmann, and K.O. Velthaus, Conf. Rec. 1997 Inter.Display Res. Conf., Toronto, pp. 297.

14. W. Tong, Y. Xin, M. Chaichimansor, J. Choi, T. Jones, W. Park, B.K. Wagner, C.J. Summers, and S.S. Sun, 4th Inter. Conf. Sci. Tech. Display Phosphors, Extended Abstracts, Bend, OR, 1998, pp. 339.

15. H.P. Maruska, T. Parodos, N.M. Kalkhoran, and W.D. Halverson, Mater. Res. Soc. Symp. Proc. 345 (1994) 269.

16. R.T. Holm, S.W. McKnight, E.D. Palik and W. Lukosz, Appl. Optics 21 (1982) 2512.