2008
DOI: 10.1063/1.2839379
|View full text |Cite
|
Sign up to set email alerts
|

Impact of incomplete set programing on the performance of phase change memory cell

Abstract: Phase change memory (PCM) cells with T-shaped structure using tungsten heater were fabricated and the cell characteristics concerning the programing pulse width were also investigated in this work. The numerical modeling shows the thermal nonuniformity over the active region due to the considerable thermal sink of tungsten heater results in the amorphous-phase residues and the incomplete set programing. The experimental results reveal the existence of residual amorphous phase and indicate that the incomplete s… Show more

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
2
2

Citation Types

0
17
0

Year Published

2009
2009
2022
2022

Publication Types

Select...
6
2

Relationship

0
8

Authors

Journals

citations
Cited by 37 publications
(17 citation statements)
references
References 5 publications
0
17
0
Order By: Relevance
“…It can be observed that the shorter the time is, the larger is the voltage. This is because during a longer heating period, the wider pulse helps to decrease the required driving voltage with respect to nearly the same input energy for crystallizing or melting the active region in the cell [26]. Note that the reversible phase change can be realized by a pulse as short as 8 ns, greatly shorter than that of GST (∼50 ns [27]), Sb 10 Se 10 Te 80 and Sb 15 Se 15 Te 70 (both about 100 ns [19]).…”
Section: Resultsmentioning
confidence: 98%
“…It can be observed that the shorter the time is, the larger is the voltage. This is because during a longer heating period, the wider pulse helps to decrease the required driving voltage with respect to nearly the same input energy for crystallizing or melting the active region in the cell [26]. Note that the reversible phase change can be realized by a pulse as short as 8 ns, greatly shorter than that of GST (∼50 ns [27]), Sb 10 Se 10 Te 80 and Sb 15 Se 15 Te 70 (both about 100 ns [19]).…”
Section: Resultsmentioning
confidence: 98%
“…4a-d could be defined as 320, 280, 260 and 260 ns, respectively. Consequently, we cannot effectively exploit the high crystallization rate of the conventional GST225, unless the specified structures, such as the modification of heater materials [4,8], are provided for the complete crystallization. The clear SET operations thanks to the perfect crystallization in moderate time can preferably do good for the retention behaviors even at a relatively hightemperature.…”
Section: Resultsmentioning
confidence: 99%
“…On the other hand, the stabilities in SET operations, such as SET resistance ðR SET Þ uniformity between devices within a single array and small variations of R SET in the same device at each operation, are also very important [1][2][3], although it has not been focused as a main issue of PRAM so much. One of the main reasons for the SET stability degradation is the imperfect crystallization process of phase-change material [4] when the SET currents ðI SET Þ having a given pulse width are applied to the PCMs. This incomplete SET operation is closely related to the existence of threshold voltage ðV th Þ required for the electronic switching of the chalcogenide alloys.…”
Section: Introductionmentioning
confidence: 99%
“…The sensing margin (defined as R reset =R set ) is $75, which is quite larger than that of the Ge 2 Sb 2 Te 5 based PCM cell ($20) operated under 50 ns voltage pulse. 13) This indicates that the Si 2 Sb 2 Te 3 based cell can achieve a more complete crystallization set state even under shorter pulse width. Moreover, the resistance drop and climb of the Si 2 Sb 2 Te 3 based cell are remarkably steeper than those of the Ge 2 Sb 2 Te 5 based cell.…”
mentioning
confidence: 92%