전산 모사를 이용한 유기물 메모리 소자의 전기적 성질에 대한 연구

Abstract

DRAM 및 플래시 메모리의 빠른 처리 속도와 대규모 저장 능력은 다수의 정보의 빠른 처리와 장시간 저장을 가능하게 주었다. 그러나 DRAM 및 플래시 메모리의 성능과 제조는 물리적, 기술적으로 점차 그 한계에 다다르고 있다. 이러한 문제를 해결하기 위해 유기물 또는 유기물과 무기물의 복합체를 기억 매체로 사용하는 유기물 메모리 소자에 대한 연구가 최근 활발히 진행되고 있다. 유기물 메모리 소자는 두 전극 사이에 단층 또는 다층의 유기물 구조가 있는 구조로서, 구조가 간단하고 제작비가 저렴하며 기계적인 유연함이 높다는 장점을 가지고 있다. 그러나, 소자의 전기적인 특성이나 기억 메커니즘에 대한 수학적 모델의 연구는 거의 진행되고 있지 않다. 본 연구에서는 유기물 내의 전하의 주입과 수송 및 기억 메커니즘에 대한 시뮬레이션 모델을 수립하고, 수립된 모델을 바탕으로 유기물 메모리 소자에서 나타나는 전류-전압 특성, 기억 특성, 부성 미분 저항 및 유기물 이동도에 대한 시뮬레이션을 수행하였다. 2장에서는 기존에 연구되던 FeRAM, MRAM, PRAM 및 멤리스터와 같은 다양한 차세대 비휘발성 메모리 소자들의 기본적인 원리와 구조를 제시하였으며, 유기물 메모리 소자에 대한 기본적인 특성을 서술하였다. 3장에서는 구동 전압 및 트랩 특성에 따른 유기물 메모리 소자의 기억 특성에 대한 시뮬레이션을 수행하였다. 메모리 소자의 기억 특성은 전극과 유기물 계면에 존재하는 트랩된 전자에 의해 나타난다. 쓰기 전압이 증가할수록 읽기 동작에서의 ON/OFF 비율이 증가한다. 쓰기 동작에서 트랩된 전자에 의해 발생하는 내부 전계가 읽기 동작에서 주입되는 정공의 주입 효율을 증가시켜 정공 전류가 증가하고, 이것이 메모리 소자에서 기억 특성이 나타나게 한다. 4장에서는 공간 전하 제한 전류 모델을 사용하여 유기물 메모리 소자의 기억 상태가 변화할 때 소자의 전도도가 낮은 상태에서 높은 상태로 급격이 전이되는 현상을 시뮬레이션 하였다. 트랩이 부분적으로 점유되면 소자의 상태는 낮은 전도도에서 높은 전도도로 급격하게 변화한다. 모든 트랩이 전자에 의해 점유되면, 더 이상의 전도도의 변화는 없다. 계산치와 SnO2 나노 입자를 사용하여 실제로 제작한 유기물 소자의 I-V 측정치와 계산치를 비교한 결과, 계산식에 병렬 저항을 삽입함으로써 모든 영역에서 측정치와 계산치가 일치하였다. 5장에서는 단층 구조의 유기물 메모리 소자에서 발생하는 부성 미분 저항 현상을 시뮬레이션 하였다. 트랩의 농도가 높아질수록 ON/OFF 비율이 증가하고 트랩 농도가 특정 수치를 초과하면 부성 미분 저항 현상이 발생한다. 트랩의 농도가 증가할수록 비어있는 트랩의 농도가 증가하여 전압이 증가할수록 전자의 포획율이 지속적으로 커진다. 그 결과 트랩된 전자에 의한 내부 전계의 증가율이 외부 전압에 의한 외부 전계의 증가율보다 커지게 되어 부성 미분 저항 현상이 나타나게 된다. 6장에서는 유기물 내부와 계면에서의 전하 이동도를 몬테카를로 방법을 사용하여 시뮬레이션 하였다. 전계와 온도에 따른 이동도의 변화는 유기물 내부와 계면에서 서로 다르게 나타난다. 전계에 따른 이동도 특성은 유기물 내부와 계면에서 모두 Poole-Frenkel 특성이 나타난다. 이동도의 온도에 대한 의존성은 유기물 내부에서는 1/T2으로 나타나며, 계면에서는 1/T의 관계를 따른다. 이러한 원인은 깊은 준위의 트랩 밀도가 유기물 내부와 계면에서의 이동도 특성의 차이를 발생시킨다. 앞으로의 시뮬레이션에 대한 연구는 다양한 종류의 시뮬레이션 모델을 하나의 단일 모델로 통합하는 것이 될 것이다. 표동-확산 모델을 기반으로 하여 낮은 전압에서의 저항성 전류와 보다 정확한 유기물 내의 이동도 모델을 접목할 것이다.|The DRAM and flash memory with a high processing speed and storage density enabled to be possible to process and store the mass information. However, since the device performance and fabrication process of the DRAM and flash memory face to the physical and technical limits. To resolve these problems based on the silicon, the investigation of the organic memory devices have been conducted, which are fabricated by the organic materials or organic/inorganic composites as the storage media. The basic structure of the organic memory device consists of one or multi layer between two parallel electrodes, which has many potential advantages such as the simple structure, cheap fabrication cost, and good mechanical flexibility. However, theoretical studies using mathematical models on the electrical characteristic and memory mechanism are less. This study was performed to establish the simulation models by using various suggested mathematical models on the carrier injection, carrier transport, and memory mechanisms in the organic materials, and investigate the theoretical calculations of the I-V curve, memory effect and negative differential resistance effect on various materials parameters in the materials and structures of the organic memory by the established simulation models. In the chapter 2, studies on various next-generation memory devices at the present and the properties of the organic memory are introduced and the definition of the organic memory device, the carrier transport, and memory mechanisms are described. In the chapter 3, the mathematical model is constructed for simulating the memory effects on the operation voltage and trap characteristics. The memory effect appears by the trapped electrons near the interface. As the writing voltage increases, the ON/OFF ratio at the reading operation increases. Because of the internal electric field induced from the trapped electrons, the hole injection efficiency and the hole mobility increases, resulting that the ON state current is relatively larger than the OFF state current. In the chapter 4, the state transition from low conductivity state to the high conductivity state is simulated by using the space charge limited current model. When all traps are completely occupied by electrons, the device state is transited from low to high conductivity, resulting in an abrupt increase in the current. The calculated I-V curves are compared with the experimental I-V curves from the fabricated devices with the different concentrations of SnO2 nanoparticles. Adding the parallel resistor, the calculated and experimental results are well agreed with each other. In the chapter 5, the NDR behavior of organic memory device with a single organic layer is simulated. When the trap density increases, the current bistability increases, and the NDR behavior appears in I-V curve. Because the unoccupied trap density increases with increasing the trap density, the electron trapping rate increases with increasing applied voltage. The increasing rate of the internal electric field from the trapped electrons becomes larger than that of the external electric field from the external applied voltage, resulting in the appearance of the NDR behavior. In chapter 6, the organic mobility behaviors in the bulk and in the interface are investigated by using the Monte-Carlo simulation. The dependence of mobilities on the electric field and the temperature shows a difference of the mobility behavior between in the bulk and in the interface. The mobility behaviors on the electric field both in the bulk and in the interface follow in the Poole-Frenkel behavior. The dependence on the temperature showed the 1/T2 relation in the bulk and the 1/T relation in the interface. Therefore, the density of the deep sites significantly affected the difference in the mobility behavior between the in the bulk and in the interface. In future, various simulation models will be unified to the single model. The new unified simulation model based on the drift-diffusion model will be developed, where the ohmic current model in the low voltage and organic mobility model is inserted in.; The DRAM and flash memory with a high processing speed and storage density enabled to be possible to process and store the mass information. However, since the device performance and fabrication process of the DRAM and flash memory face to the physical and technical limits. To resolve these problems based on the silicon, the investigation of the organic memory devices have been conducted, which are fabricated by the organic materials or organic/inorganic composites as the storage media. The basic structure of the organic memory device consists of one or multi layer between two parallel electrodes, which has many potential advantages such as the simple structure, cheap fabrication cost, and good mechanical flexibility. However, theoretical studies using mathematical models on the electrical characteristic and memory mechanism are less. This study was performed to establish the simulation models by using various suggested mathematical models on the carrier injection, carrier transport, and memory mechanisms in the organic materials, and investigate the theoretical calculations of the I-V curve, memory effect and negative differential resistance effect on various materials parameters in the materials and structures of the organic memory by the established simulation models. In the chapter 2, studies on various next-generation memory devices at the present and the properties of the organic memory are introduced and the definition of the organic memory device, the carrier transport, and memory mechanisms are described. In the chapter 3, the mathematical model is constructed for simulating the memory effects on the operation voltage and trap characteristics. The memory effect appears by the trapped electrons near the interface. As the writing voltage increases, the ON/OFF ratio at the reading operation increases. Because of the internal electric field induced from the trapped electrons, the hole injection efficiency and the hole mobility increases, resulting that the ON state current is relatively larger than the OFF state current. In the chapter 4, the state transition from low conductivity state to the high conductivity state is simulated by using the space charge limited current model. When all traps are completely occupied by electrons, the device state is transited from low to high conductivity, resulting in an abrupt increase in the current. The calculated I-V curves are compared with the experimental I-V curves from the fabricated devices with the different concentrations of SnO2 nanoparticles. Adding the parallel resistor, the calculated and experimental results are well agreed with each other. In the chapter 5, the NDR behavior of organic memory device with a single organic layer is simulated. When the trap density increases, the current bistability increases, and the NDR behavior appears in I-V curve. Because the unoccupied trap density increases with increasing the trap density, the electron trapping rate increases with increasing applied voltage. The increasing rate of the internal electric field from the trapped electrons becomes larger than that of the external electric field from the external applied voltage, resulting in the appearance of the NDR behavior. In chapter 6, the organic mobility behaviors in the bulk and in the interface are investigated by using the Monte-Carlo simulation. The dependence of mobilities on the electric field and the temperature shows a difference of the mobility behavior between in the bulk and in the interface. The mobility behaviors on the electric field both in the bulk and in the interface follow in the Poole-Frenkel behavior. The dependence on the temperature showed the 1/T2 relation in the bulk and the 1/T relation in the interface. Therefore, the density of the deep sites significantly affected the difference in the mobility behavior between the in the bulk and in the interface. In future, various simulation models will be unified to the single model. The new unified simulation model based on the drift-diffusion model will be developed, where the ohmic current model in the low voltage and organic mobility model is inserted in.