Colloidal Quantum Dots Nanocomposites Based Resistive Switching for Low-Power Resistive Random Access Memory
Loading...
Date
2022-12
Researcher
Bera, Jayanta
Supervisor
Sahu, Satyajit
Journal Title
Journal ISSN
Volume Title
Publisher
Indian Institute of Technology Jodhpur
Abstract
Memory is one of the most essential parts in human body or machine. Conventional memory technologies such as HDD, USB Flash drive, SSD etc. which are being used as the data storage devices over past few decades, are failed to store the huge amount of data generated by explosion of digital information and internet of things (IoT). The silicon-based technologies are facing several theoretical and practical limitations upon downscaling. Also, the Von Neumann architecture used in the current computers, has separate processing unit and storage unit. As the processing unit and storage units are separated and connected by buses, the data transfer speed is limited and this architecture consumes lot of power. Therefore, it is essential to design universal memory with high density data storage, fast operation speed, capacity of in memory computing and low power consumption. Emerging memory technologies such as phase change memory (PCM), ferroelectric random-access memory (FERAM), magnetic random access memory (MRAM) and resistive random access memory (RRAM) have drawn much attention to be used as universal memory. RRAM is one of the most promising for universal memory candidates owing to its simple architecture, high stacking density, fast operation speed, capability of in memory computing, long retention time, high endurance cycles and low operational power. The working principle of RRAM is based on the change in the resistance of the device modulated by electrical stimulation. High resistance state is the off state which can be considered as bit 0 and low resistance state is the on state which can be considered as bit 1 in RRAM. Depending on the number of different resistance state, multibit can be stored in RRAM.
Semiconducting quantum dots (QDs) embedded in polymer matrix or sandwiched between thin oxide layers can be used as an active material for high performance RRAM device. QDs having the particle size in the range between 2 to 10 nm show quantum confinement effect. Various surface trap states are present in the QDs. These surface trap states can be modified by encapsulating the surface with different ligands. In the composite of QDs with polymers, QDs act as charge trapping centers. Colloidal CdSe QD by the hot injection method was synthesized and these QDs were blended with poly(4-vinylpyridine) (PVP) polymer to prepare the active layer of a RRAM device. For device fabrication, the CdSe QDs-PVP nanocomposite was spin coated on an ITO substrate and finally Al top electrode was deposited using a thermal evaporation unit. The Al/CdSe QDs-PVP/ITO device exhibited excellent bipolar resistive switching (RS) behavior with maximum on-off ratio (ION/IOFF) of 105, long retention time (> 3700 s), high endurance. The SET and RESET voltage of this device was +1.6 V and -1.7 V respectively.
Another device using CdSe QDs-PVP nanocomposites as an active material with slightly different ratio from previous one was fabricated with Al as both top and bottom electrodes. This Al/CdSe QDs-PVP/Al device also showed excellent resistive switching with less SET and RESET voltage as compared to previous one. The SET and RESET voltages of Al/CdSe QDs-PVP/Al device were +0.6 V and -0.5 V respectively. The ultimate power consumption of this device is 4.16 ?W in the on state and 80 pW in the off state, which are sufficiently low to be considered as a low power memory device. Also, the Al/CdSe QDs-PVP/Al device has data retention time larger than 35,000 s which is much better than previous Al/CdSe QDs-PVP/ITO device. The resistive switching in CdSe QDs-PVP nanocomposites-based device is ascribed to charge trapping and detrapping in QDs while the polymers act as charge blocking layer.
Though CdSe QDs based RRAM shows excellent resistive switching behavior with low power consumption, CdSe is toxic and found to be human carcinogens. Therefore, in search of less toxic and non-carcinogens materials, we have synthesized colloidal Molybdenum disulfide (MoS2) QDs. The MoS2 QDs were blended with PVP and used as active material in Al/MoS2 QDs-PVP/ITO device. The Al/MoS2 QDs-PVP/ITO device also exhibited excellent RS with high value of ION/IOFF (~ 105), long retention time (> 25,000 s), high endurance over 250 cycles and very fast response time (~ 28 ns). The operating voltage of the device was less than states. In fact, our study further demonstrates the presence of nonlocality in mixed and separable
states where even the Bell or Bell-type inequalities fail to capture nonlocal correlations. The results
successfully address issues of bipartite vs tripartite vs multiqubit nonlocality and further identify
all pure multiqubit entangled states for the complete range of state parameters. The discussion
is extended to quantify nonlocality in different classes of four and five qubit entangled pure
states. Furthermore, the analytical results obtained in this Thesis are in complete agreement with
numerical results.
Based on our studies for bipartite and multiqubit systems, we readdress the usefulness
of partially entangled multiqubit states for quantum information and computation. For this,
we revisit the question of analyzing efficiencies of four different sets of partially entangled
states in three qubit classes under real conditions. In the presence of noise, we show that
maximum entanglement and nonlocality in the input state do not always guarantee maximum
efficiency in a protocol. For example, our analysis suggests that efficiencies of a set of partially
entangled states are much more robust to noise than those of maximally entangled states in
Greenberger–Horne–Zeilinger (GHZ) class of states. For a set of partially entangled states in
presence of noise and weak measurements, efficiencies of communication protocols achieve the
optimal value independent of the state and decoherence parameters. We further generalize our
study to address efficiencies of (N + 2)-qubit partially entangled states for the presence of N
controllers in a noisy environment from the perspective of controllers’ authority and average
fidelity. These values are obtained by designing a generalized circuit using single and two-qubit
gates and studying different cases of two sets of partially entangled multiqubit states. Our
analysis identifies a set of partially entangled states for which the average fidelity is independent
of the state parameter and measurements performed by (N − 1) controllers thereby facilitating
the experimental set-ups to worry about a smaller number of parameters in the protocol when
dealing with a multiqubit network. Interestingly, even in the presence of noise and weak
measurement operations, we find the efficiency of a set of (N +2)-qubit partially entangled states
to be independent of the measurements performed by (N−1) controllers. The efficiencies of these
states are consistent with the analysis of nonlocal correlations using our modified operators and
quantum discord.states. In fact, our study further demonstrates the presence of nonlocality in mixed and separable
states where even the Bell or Bell-type inequalities fail to capture nonlocal correlations. The results
successfully address issues of bipartite vs tripartite vs multiqubit nonlocality and further identify
all pure multiqubit entangled states for the complete range of state parameters. The discussion
is extended to quantify nonlocality in different classes of four and five qubit entangled pure
states. Furthermore, the analytical results obtained in this Thesis are in complete agreement with
numerical results.
Based on our studies for bipartite and multiqubit systems, we readdress the usefulness
of partially entangled multiqubit states for quantum information and computation. For this,
we revisit the question of analyzing efficiencies of four different sets of partially entangled
states in three qubit classes under real conditions. In the presence of noise, we show that
maximum entanglement and nonlocality in the input state do not always guarantee maximum
efficiency in a protocol. For example, our analysis suggests that efficiencies of a set of partially
entangled states are much more robust to noise than those of maximally entangled states in
Greenberger–Horne–Zeilinger (GHZ) class of states. For a set of partially entangled states in
presence of noise and weak measurements, efficiencies of communication protocols achieve the
optimal value independent of the state and decoherence parameters. We further generalize our
study to address efficiencies of (N + 2)-qubit partially entangled states for the presence of N
controllers in a noisy environment from the perspective of controllers’ authority and average
fidelity. These values are obtained by designing a generalized circuit using single and two-qubit
gates and studying different cases of two sets of partially entangled multiqubit states. Our
analysis identifies a set of partially entangled states for which the average fidelity is independent
of the state parameter and measurements performed by (N − 1) controllers thereby facilitating
the experimental set-ups to worry about a smaller number of parameters in the protocol when
dealing with a multiqubit network. Interestingly, even in the presence of noise and weak
measurement operations, we find the efficiency of a set of (N +2)-qubit partially entangled states
to be independent of the measurements performed by (N−1) controllers. The efficiencies of these
states are consistent with the analysis of nonlocal correlations using our modified operators and
quantum discord.states. In fact, our study further demonstrates the presence of nonlocality in mixed and separable
states where even the Bell or Bell-type inequalities fail to capture nonlocal correlations. The results
successfully address issues of bipartite vs tripartite vs multiqubit nonlocality and further identify
all pure multiqubit entangled states for the complete range of state parameters. The discussion
is extended to quantify nonlocality in different classes of four and five qubit entangled pure
states. Furthermore, the analytical results obtained in this Thesis are in complete agreement with
numerical results.
Based on our studies for bipartite and multiqubit systems, we readdress the usefulness
of partially entangled multiqubit states for quantum information and computation. For this,
we revisit the question of analyzing efficiencies of four different sets of partially entangled
states in three qubit classes under real conditions. In the presence of noise, we show that
maximum entanglement and nonlocality in the input state do not always guarantee maximum
efficiency in a protocol. For example, our analysis suggests that efficiencies of a set of partially
entangled states are much more robust to noise than those of maximally entangled states in
Greenberger–Horne–Zeilinger (GHZ) class of states. For a set of partially entangled states in
presence of noise and weak measurements, efficiencies of communication protocols achieve the
optimal value independent of the state and decoherence parameters. We further generalize our
study to address efficiencies of (N + 2)-qubit partially entangled states for the presence of N
controllers in a noisy environment from the perspective of controllers’ authority and average
fidelity. These values are obtained by designing a generalized circuit using single and two-qubit
gates and studying different cases of two sets of partially entangled multiqubit states. Our
analysis identifies a set of partially entangled states for which the average fidelity is independent
of the state parameter and measurements performed by (N − 1) controllers thereby facilitating
the experimental set-ups to worry about a smaller number of parameters in the protocol when
dealing with a multiqubit network. Interestingly, even in the presence of noise and weak
measurement operations, we find the efficiency of a set of (N +2)-qubit partially entangled states
to be independent of the measurements performed by (N−1) controllers. The efficiencies of these
states are consistent with the analysis of nonlocal correlations using our modified operators and
quantum discord.states. In fact, our study further demonstrates the presence of nonlocality in mixed and separable
states where even the Bell or Bell-type inequalities fail to capture nonlocal correlations. The results
successfully address issues of bipartite vs tripartite vs multiqubit nonlocality and further identify
all pure multiqubit entangled states for the complete range of state parameters. The discussion
is extended to quantify nonlocality in different classes of four and five qubit entangled pure
states. Furthermore, the analytical results obtained in this Thesis are in complete agreement with
numerical results.
Based on our studies for bipartite and multiqubit systems, we readdress the usefulness
of partially entangled multiqubit states for quantum information and computation. For this,
we revisit the question of analyzing efficiencies of four different sets of partially entangled
states in three qubit classes under real conditions. In the presence of noise, we show that
maximum entanglement and nonlocality in the input state do not always guarantee maximum
efficiency in a protocol. For example, our analysis suggests that efficiencies of a set of partially
entangled states are much more robust to noise than those of maximally entangled states in
Greenberger–Horne–Zeilinger (GHZ) class of states. For a set of partially entangled states in
presence of noise and weak measurements, efficiencies of communication protocols achieve the
optimal value independent of the state and decoherence parameters. We further generalize our
study to address efficiencies of (N + 2)-qubit partially entangled states for the presence of N
controllers in a noisy environment from the perspective of controllers’ authority and average
fidelity. These values are obtained by designing a generalized circuit using single and two-qubit
gates and studying different cases of two sets of partially entangled multiqubit states. Our
analysis identifies a set of partially entangled states for which the average fidelity is independent
of the state parameter and measurements performed by (N − 1) controllers thereby facilitating
the experimental set-ups to worry about a smaller number of parameters in the protocol when
dealing with a multiqubit network. Interestingly, even in the presence of noise and weak
measurement operations, we find the efficiency of a set of (N +2)-qubit partially entangled states
to be independent of the measurements performed by (N−1) controllers. The efficiencies of these
states are consistent with the analysis of nonlocal correlations using our modified operators and
quantum discord.states. In fact, our study further demonstrates the presence of nonlocality in mixed and separable
states where even the Bell or Bell-type inequalities fail to capture nonlocal correlations. The results
successfully address issues of bipartite vs tripartite vs multiqubit nonlocality and further identify
all pure multiqubit entangled states for the complete range of state parameters. The discussion
is extended to quantify nonlocality in different classes of four and five qubit entangled pure
states. Furthermore, the analytical results obtained in this Thesis are in complete agreement with
numerical results.
Based on our studies for bipartite and multiqubit systems, we readdress the usefulness
of partially entangled multiqubit states for quantum information and computation. For this,
we revisit the question of analyzing efficiencies of four different sets of partially entangled
states in three qubit classes under real conditions. In the presence of noise, we show that
maximum entanglement and nonlocality in the input state do not always guarantee maximum
efficiency in a protocol. For example, our analysis suggests that efficiencies of a set of partially
entangled states are much more robust to noise than those of maximally entangled states in
Greenberger–Horne–Zeilinger (GHZ) class of states. For a set of partially entangled states in
presence of noise and weak measurements, efficiencies of communication protocols achieve the
optimal value independent of the state and decoherence parameters. We further generalize our
study to address efficiencies of (N + 2)-qubit partially entangled states for the presence of N
controllers in a noisy environment from the perspective of controllers’ authority and average
fidelity. These values are obtained by designing a generalized circuit using single and two-qubit
gates and studying different cases of two sets of partially entangled multiqubit states. Our
analysis identifies a set of partially entangled states for which the average fidelity is independent
of the state parameter and measurements performed by (N − 1) controllers thereby facilitating
the experimental set-ups to worry about a smaller number of parameters in the protocol when
dealing with a multiqubit network. Interestingly, even in the presence of noise and weak
measurement operations, we find the efficiency of a set of (N +2)-qubit partially entangled states
to be independent of the measurements performed by (N−1) controllers. The efficiencies of these
states are consistent with the analysis of nonlocal correlations using our modified operators and
quantum discord ±1.5 V. Furthermore, this MoS2 QDs-PVP based device shows stable resistive switching when heated even at 130°C. The excellent temperature stability of the fabricated MoS2 QDs-PVP RRAM device reveals that it can be effectively used in extremely hot weather conditions without degradation. These studies reveal that CdSe and MoS2 QDs embedded in PVP matrix can have great potential to be used as active materials for stable, high performance and low power RRAM dev
Description
Keywords
Citation
Bera, Jayanta. (2022). Colloidal Quantum Dots Nanocomposites Based Resistive Switching for Low-Power Resistive Random Access Memory (Doctor's thesis). Indian Institute of Technology Jodhpur, Jodhpur.