Fabrication of visible a-SiC:H-based p-i-n type thin-film light-emitting diodes (TFLEDs) using a photo-CVD method is proposed. a-SiC:H films, the important luminescently active i layer of the TFLEDs, are prepared by a direct photo-CVD method from a gas mixture of disilane and ethylene. a-SiC:H films have a bandgap enough to be applicable to the luminescently active i layer of the TFLEDs and also show good photoluminescence characteristics. In these diodes, p-μc-Si:H and n-μc-Si:H films were used as hole and electron injectors into the luminescently active i layer, respectively. The fabricated visible a-SiC:H-based p-i-n type TFLED has a vertical structure of glass/$SnO_2$ (F-doped)/p-μc-Si:H (2.3 eV)/i-a-SiC:H (3.0 eV)/n-μc-Si:H (2.4 eV)/Al. It has a threshold voltage of 6V and electroluminescence (EL) peak wavelength of 700 nm emitting reddish light as detected by the naked eyes.
To enhance the efficiency of hole injection into the luminescently active i layer, a HTI layer was inserted at the p/i interface. It is found that inserting the HTI layer increases the efficiency of hole injection, improving the performance of the TFLEDs. The obtained optimum thickness of the HTI layer was 3.0 nm. Then, the effect of thickness and bandgap of p-μc-Si:H layer on the performance of TFLEDs was investigated. Increasing the bandgap of p-layer decreases the height of notch barrier between p-μc-Si:H and HTI layers, resulting in injection of the more amount of holes to be injected into the luminescently active i layer.
Then, to improve the device performance by decreasing the amount of carriers trapped by defect states at the HTI/i interface, the HTI layer was graded. Such a grading was obtained by decreasing the flow rate of hydrocarbon gas (i.e. ethylene) from a predetermined value to zero during deposition of the layer. The visible a-SiC:H-based p-i-n type TFLEDs with a graded-gap HTI layer showed better device characteristics than those with a nongraded-gap HTI layer.
However, although the performance of visible a-SiC:H-based p-i-n type TFLEDs was improved by various band structure tailoring techniques as mentioned above, their brightness was still not enough for a practical application. Thus, we proposed a new technique, hydrogen passivation process, and applied it ex-situtively and in-situtively to visible a-SiC:H-based p-i-n type TFLEDs. An ex-situ hydrogen passivation process was carried out by a plasma apparatus, ICP equipment, after the TFLED fabrication. The ex-situ hydrogenated TFLEDs showed drastically-improved device performance; The threshold voltage decreased by about 5 V, the EL intensity increased by a factor of about three, and the EL peak shifted towards a shorter wavelength, from 704.5 nm to 680 nm, resulting in increase of the brightness by 35 times from 1 cd/㎡ at an injection current density of 360 mA/㎠ to 35 cd/㎡ at an injection current density of 1 A/㎠. Also, the process time dependence of hydrogen passivation effects was investigated. It was found that increasing the hydrogen passivation process time beyond a saturation point only degraded the device performance; that is, the threshold voltage slightly increased, the EL peak shifted towards a longer wavelength, and the brightness degraded.
Then, an in-situ hydrogen passivation process using a photo-CVD technique during the TFLED fabrication process was performed: it was carried out at the base of a p-μc-Si:H layer with various thickness of an i-a-SiC:H layers (0 nm, 10 nm, 20 nm, and 30 nm). The in-situ hydrogenated TFLEDs showed much more drastically-improved device performance than that of the ex-situ hydrogenated TFLED. In particular, the brightness increased up to 80 cd/f㎡ at an injection current density of 1 A/㎠.
Finally, we combined two hydrogen passivation processes, in-situ and ex-situ processes, to improve the device performance. In this combined process (two-step hydrogen passivation process), the in-situ hydrogen process was followed by the ex-situ one. Using the two-step hydrogen passivation process, the device performance was dramatically improved; that is, the threshold voltage was decreased by about 2 V, the EL peak shifted towards a much shorter wavelength, and the brightness increased to 128 cd/㎡.
As mentioned above, the device performance was remarkably improved by various hydrogen passivation processes. This improvement of device performance was believed to be caused by passivation of defect states by hydrogen atoms. In order to clarify possible mechanisms by which the device performance can be improved, ESR spectrum was measured on an as-deposited and hydrogenated a-SiC:H films at room temperature. It was found that the spin density (i.e. density of dangling bonds) within a-SiC:H film was decreased by various hydrogen passivation processes. Thus, it was concluded that a hydrogen passivation process reduced the density of defect states at all interfaces (i.e. p/i and i/n interfaces) as well as in the luminescently active i-a-SiC:H layer, resulting in dramatic improvement in the device performance. These results should encourage the use of hydrogen passivation process to improve the performance of visible p-i-n type TFLEDs.
광-CVD법을 이용한 p-i-n형 가시광 발광 박막 다이오드의 제작이 제안되었다. 박막 발광 다이오드에서 가장 중요한 발광 능동 I층인 비정질 탄화규소 박막은 $Si_2H_6$와 $C_2H_4$의 혼합가스를 이용하여 직접 여기 광-CVD법에 의해 형성되었다. 형성된 비정질 탄화규소 박막은 박막 발광 다이오드의 발광 능동 i층으로 응용될 수 있을 정도의 광학적 밴드갭과 양호한 PL(photoluminescence) 특성을 보였다. 제작된 박막 발광 다이오드에서, p형 미결정 실리콘 박막과 n형 미결정 실리콘 박막이 발광 능동 I층으로의 정공 및 전자 주입층으로 각각 이용되었다. 제작된 p-i-n형 가시광 발광 박막 다이오드의 단면구조는 glass/$SnO_2$ (F-doped)/p-μc-Si:H (2.3 eV)/i-a-SiC:H (3.0 eV)/n-μc-Si:H (2.4 eV)/Al이었다.
발광 능동 i층으로의 정공 주입 효율을 증가시키기 위하여, HTI (Hot-carrier Tunneling Injection)층이 p/i계면에 삽입되었으며 HTI층의 최적 두께는 약 3.0 nm였다. 또한 HTI층을 grading하여 HTI층과 i층 계면에서 결함에 의해 포획되는 캐리어의 양을 줄임으로써 소자의 특성이 향상되었다. p층의 밴드갭 및 두께가 소자의 특성에 미치는 효과도 검토되었으며, 밴드갭이 클 수록 그리고 막이 두꺼울 수록 소자의 특성이 향상된다는 것을 알았다.
그러나, 비록 p-i-n형 가시광 발광 박막 다이오드의 성능이 위에서 언급된 다양한 band structure tailoring을 통해 향상되었다 할지라도, 제작된 소자의 특성, 특히 휘도, 은 박막 발광 다이오드를 실제로 응용하기 위해서는 충분하지 못했다. 그리하여, 본 연구에서는 새로운 기법인 hydrogen passivation process을 제안하였으며, p-i-n형 가시광 발광 박막 다이오드에 여러가지 형태로 적용하였다. Ex-situ hydrogen passivation process는 박막 발광 다이오드를 제작한 후 플라즈마 장치를 이용하여 수행되었으며 소자의 성능을 현저하게 향상시켰다. 즉, 문턱 전압은 약 5V정도 감소했으며, EL(electroluminescence)의 세기는 수배 정도 증가했으며, EL피크는 더 짧은 파장으로 이동했으며(704.5 nm에서 630 nm), 결과적으로 휘도는 1cd/㎡에서 약 35cd/㎡로 뚜렷하게 증가했다. 또한, hydrogen passivation 효과의 process time 의존성도 조사되었으며, hydrogen passivation process time을 증가시킴에 따라 소자의 성능이 열화되는 것이 밝혀졌다.
In-situ hydrogen passivation process는 박막 발광 다이오드 제작 공정 중에 광-CVD법에 의해 p층 및 발광 능동 I층에서 수행되었다. In-situ hydrogen passivation process는 ex-situ hydrogen passivation process보다 훨씬 효과적이었으며, 이 방법에 의해 소자의 휘도는 약 80 cd/㎡까지 증가되었다.
그런다음, in-situ와 ex-situ hydrogen passivation process를 조합한 two-step hydrogen passivation process가 수행되었다. 이 방법에서는 in-situ process가 ex-situ process보다 먼저 수행되었다. 이 two-step process에 의해 소자의 성능은 급격하게 향상되었으며, 특히 소자의 휘도는 약 128 cd/㎡까지 증가되었다.
위에서 언급된 바 대로, 소자의 성능은 여러가지 hydrogen passivation process에 의해 현저하게 향상되었다. 이와 같은 소자의 성능 향상에 관여한 mechanisms을 규명하기 위해 ESR(electron spin resonance)측정을 수행하였으며, hydrogen atoms에 의해 비정질 탄화규소 박막내에 존재하는 spin density, 즉 dangling bonds의 양이 감소된다는 것이 밝혀졌다. 따라서, hydrogen passivation process에 의해 모든 계면(p/i 및 i/n 계면) 그리고 발광 능동 i층에 존재하는 결함 밀도가 감소되어 박막 발광 다이오드의 성능이 급격히 향상되었다는 결론을 이끌어낼 수 있었다. 끝으로, 이러한 결과는 p-i-n형 가시광 발광 박막 다이오드의 성능을 향상시키기 위해 hydrogen passivation process의 응용을 적극 권장하는 계기가 될 것이다.
