Impedance Spectroscopy of novel Access Devices based on ... .fr

Figure 1.3: 3D multi-layer architecture : solution for huge density SCM. 1PCM exploits ..... First, a periodic perturbation is .... experimental values and DFT calculation in the table 2.1. ..... SPICE simulations have been done during this internship.
7MB taille 3 téléchargements 287 vues
IBM Research - Almaden

Master Nanotech

650 Harry Road

Phelma, Grenoble INP

San Jose, CA 95120

Academic year: 2010-2011

Impedance Spectroscopy of novel Access Devices based on Mixed Ionic Electronic Conduction

Author:

Supervisors:

Benjamin Meunier [email protected]

Georey Burr [email protected]

April-August 2012

1

Contents Introduction 1 Global presentation of the master thesis 1.1

IBM Almaden Research Center

1.2

Motivation . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

7

1.2.1

Storage-Class Memory (SCM) . . . . . . . . . . . . . .

7

1.2.2

Flash limitation . . . . . . . . . . . . . . . . . . . . . .

7

1.2.3

3D Multi-layer SCM

1.3

. . . . . . . . . . . . . . . . .

4 6

. . . . . . . . . . . . . . . . . . .

8

State-of-Art . . . . . . . . . . . . . . . . . . . . . . . . . . . .

9

1.3.1

General principle

. . . . . . . . . . . . . . . . . . . .

9

1.3.2

Mathematical model . . . . . . . . . . . . . . . . . . .

10

1.3.3

Previous results from IBM . . . . . . . . . . . . . . . .

12

2 Description of the experiments 2.1

2.2

IV characteristics of MIEC . . . . . . . . . . . . . . . . . . . .

3.2

3.3

16 16

2.1.1

Conductive AFM measurement . . . . . . . . . . . . .

16

2.1.2

Description of the results

. . . . . . . . . . . . . . . .

18

. . . . . . . . . . . . . . . . . . . . .

22

Impedance spectroscopy 2.2.1

Principle of IS

. . . . . . . . . . . . . . . . . . . . . .

22

2.2.2

Limitation . . . . . . . . . . . . . . . . . . . . . . . . .

24

2.2.3

Sample sizing . . . . . . . . . . . . . . . . . . . . . . .

25

3 IS Results and discussion 3.1

6

Theoritical model for MIEC Impedance

. . . . . . . . . . . .

3.1.1

Band Diagram-based model (model 1)

3.1.2

Electrolyte like model (model 2)

3.1.3

First MIEC IS and tting

Started experiments

. . . . . . . . .

32

32

32

. . . . . . . . . . . .

33

. . . . . . . . . . . . . . . .

36

. . . . . . . . . . . . . . . . . . . . . . .

38

3.2.1

Eect of the Bias . . . . . . . . . . . . . . . . . . . . .

38

3.2.2

Eect of the thickness

. . . . . . . . . . . . . . . . . .

40

Next steps . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

42

Conclusion

44 2

Appendix A.1

Mathematical complement . . . . . . . . . . . . . . . . . . . .

47 48

A.1.1

Defect model p

1010

cycles at low

current) of such MIEC-based access devices have been well studied. Despite these accomplishments, the theoretical understanding of electronic conduction in this MIEC material remains incomplete. In this work, Impedance Spectroscopy (IS) is used to study the novel-diode MIEC material and the coupled electronic and ionic processes responsible for its attractive electrical characteristics. First, the same small samples used for C-AFM-based measurements of the IV-characteristics are tested using Impedance Spectroscopy. Such samples are found to be unsuitable for IS, because the signicant tip capacitance overwhelms the much smaller signal due to the sample impedance. Hence, a new sample structure with much larger areas has been developed. In addition, a simple and exible non-AFM-based probe-station has been established, and a specic IS measurement procedure devised. Various issues that can interfere with the spectroscopy are identied and corrected, using purely-resistive and purely-capacitive test samples. Two theoretical circuit models for the MIEC device impedance are then developed to explain the measured frequency response of the tested samples. Impedance spectroscopy is carried out on MIEC samples and the resulting curves are matched against these theoretical models.

Intriguing

and promising initial results are shown from studies into the eects of the voltage bias, the thickness of the samples, the metal used as the top electrode, and the role of post-fabrication annealing. However, signicant further research will need to be performed before IS can contribute to the theoretical understanding to the novel-diode MIEC material. Ideas for future experimental improvements include the creation of both large and small samples on the same substrate (to ensure that the familiar IV characteristics are still being obtained), the implemention of vertically symmetric cells (to investigate the observed voltage-polarity-asymmetric behavior), the extension of IS to higher frequencies, and further renement and automation of the tting procedures used to extract theoretical circuit parameters from the IS data.

55