AlGaN/GaN HEMTs : Différence entre versions
m (a déplacé InitRech 2015/2016, sujet 21 vers AlGaN/GaN HEMTs) |
(→Summary) |
||
| Ligne 1 : | Ligne 1 : | ||
__TOC__ | __TOC__ | ||
= Summary = | = Summary = | ||
| + | <p> This article is talking about '''AlGaN/GaN high electron mobility transistor (HEMT)'''. It is an heretostructure field effect transistor (HFET).<br /> | ||
| + | This kind of transistor allows to the electron a best mobility. In fact, a heterojunction with a '''highly n-doped bandgap''' (here AlGaN) and '''a undoped bandgap''' (here GaN) allow to electrons to go faster on the other side. The layer between this 2 sides is called '''2-DEG (two-dimensional electron gas)'''.<br /><br /> | ||
| + | Authors introduces first the subject, their target, and why fabricate this kind of transistor. It's a molecular beam epitaxy. A 21-nm AlGaN barrier is grown on a silicon substrate. In a second time, this barrier is processed on a rigid Si substrate by lithography. Only after this 2 steps, the transistor is transferred on a flexible substrate.<br /><br /> | ||
| + | After an explanation of DC characteristics of the transistor on a Silicium (Si) subtrate, they explain '''the effect of the transfer on a flexible substrate''' on the DC transistor characteristics. They try to explain all changes on the Ids-Vds characteristics, on C-V profile, on the Nd-Na profile, in order to show which parameters are modified, and what are the consequences of this modifications.<br /> | ||
| + | The first observation is on the maximum drain current density. It drops down when the transistor is transferred on a flexible substrate (930 mA/mm before, 450 mA/mm after the transfer). | ||
| + | The second observation is the behavior of the transfer characteristics. It does not follow the ideal transfer characteristics.<br /> | ||
| + | This 2 observations are '''due to the poor thermal conductivity of the polymer''' which form the flexible tape. In order to confirm this hypothesis, they decided to plot the C-V profile. They show that no deviation is observed in C-V profile in terms of frequency dispersion and charges density. They confirm that the '''2-DEG density is not affected and assume that no additional defects are created during the transfer'''.<br /><br /> | ||
| + | After this study, they try to show '''the effect of the bending''' on the DC characteristics. They explain that the characteristics are not the same if the flexible tape is flat or if it is bent. They realize the bending with a semicylindrical chuck with 15 mm raduis of curvature. With this chuck, strain is in the gate direction. The effect on the I-V characteristic is light. And due to the difference between the thermal conductivity between the 2 chuck they use for analysis, they do analysis with a low drain-source voltage. They show that the on-resistance (Ron) is decreased, and the sheet resistance (Rs - 11% decrease) also. It is due to the increase of 2-DEG density (in 2-DEG, the parameter which translate the mobility of the electron (μn) increased, and sheet resistance is directly close to this parameter)<br /></p> | ||
| + | |||
= Main contribution = | = Main contribution = | ||
= Applications = | = Applications = | ||
Version du 1 juin 2016 à 12:24
Sommaire
Summary
This article is talking about AlGaN/GaN high electron mobility transistor (HEMT). It is an heretostructure field effect transistor (HFET).
This kind of transistor allows to the electron a best mobility. In fact, a heterojunction with a highly n-doped bandgap (here AlGaN) and a undoped bandgap (here GaN) allow to electrons to go faster on the other side. The layer between this 2 sides is called 2-DEG (two-dimensional electron gas).
Authors introduces first the subject, their target, and why fabricate this kind of transistor. It's a molecular beam epitaxy. A 21-nm AlGaN barrier is grown on a silicon substrate. In a second time, this barrier is processed on a rigid Si substrate by lithography. Only after this 2 steps, the transistor is transferred on a flexible substrate.
After an explanation of DC characteristics of the transistor on a Silicium (Si) subtrate, they explain the effect of the transfer on a flexible substrate on the DC transistor characteristics. They try to explain all changes on the Ids-Vds characteristics, on C-V profile, on the Nd-Na profile, in order to show which parameters are modified, and what are the consequences of this modifications.
The first observation is on the maximum drain current density. It drops down when the transistor is transferred on a flexible substrate (930 mA/mm before, 450 mA/mm after the transfer).
The second observation is the behavior of the transfer characteristics. It does not follow the ideal transfer characteristics.
This 2 observations are due to the poor thermal conductivity of the polymer which form the flexible tape. In order to confirm this hypothesis, they decided to plot the C-V profile. They show that no deviation is observed in C-V profile in terms of frequency dispersion and charges density. They confirm that the 2-DEG density is not affected and assume that no additional defects are created during the transfer.
After this study, they try to show the effect of the bending on the DC characteristics. They explain that the characteristics are not the same if the flexible tape is flat or if it is bent. They realize the bending with a semicylindrical chuck with 15 mm raduis of curvature. With this chuck, strain is in the gate direction. The effect on the I-V characteristic is light. And due to the difference between the thermal conductivity between the 2 chuck they use for analysis, they do analysis with a low drain-source voltage. They show that the on-resistance (Ron) is decreased, and the sheet resistance (Rs - 11% decrease) also. It is due to the increase of 2-DEG density (in 2-DEG, the parameter which translate the mobility of the electron (μn) increased, and sheet resistance is directly close to this parameter)
