[حفيده نيوتن]

السلام عليكم ورحمة الله وبركاته ارجو ممن لديه خلفيه بالترجمه ان يساعدني تكفون لاتتخلوا عني
1 Chapter 1: Introduction
The fast growing field of microelectronics and information processing has been based on
fabricating devices out of conductors and doped semiconductors, where the charge of the
electron is exploited. Generation to generation improvements in the performance of
microelectronic devices is currently accomplished by shrinking the size of the elements in
the chips. However, this approach will soon hit a barrier due to issues related to
fabrication of shallow channel junctions and heat management [1]. Similarly, information
storage, the second vital piece of the microelectronic revolution, depends on magnetic
storage devices that exploit the spin component of the electron [2,3]. Here, continued
performance gains are bumping up against the superparamagnetic limit associated with
nanosized features. An alternative approach for improving the performance of future
devices would be to make the individual components multifunctional. The fabrication of
devices where both of the charge and spin functionalities of the carrier were combined
could revolutionize the technology by providing for new device designs and architectures
that could potentially boost performance, reduce power consumption, and introduce new
features (e.g. instantaneous boot up and data retention in the power-down state). This
branch of electronics, which involves utilizing both the spin and charge of the electron, is
commonly referred to as spintronics [2-4]. Therefore, in pursuit of more efficient, higher
performance, and lower power-consuming devices like FETs, LEDS, MRAM, etc
research is now being conducted that is focused on the complementary use of the charge
and spin components of the electron as it might be applied to information storage and
processing [4-7]. Early work has been focused on finding suitable host materials that
2
display ferromagnetic properties above room temperature, and identifying workable
approaches to the efficient injection, transport, detection and manipulation of spinpolarized
carriers [8].
Since conventional semiconductors like Si and GaAs are diamagnetic and posses very
small g-factor (a measure of the interaction strength with an applied magnetic field), it
became necessary to search for ways to incorporate magnetic properties with
semiconducting properties. The use of ferromagnetic metals like Fe,Ni, etc as contacts for
injecting spin polarized electrons into conventional semiconductors like Si and GaAs has
been extensively studied [5,9,10]. Spin polarized carrier injection proved to be very
inefficient in these structures due to the formation of detrimental interfacial layers and
scattering of carriers due to impedance mismatch [11]. Therefore, research gradually
shifted over to finding a magnetic semiconductor that could act as a source for the
efficient injection of spin-polarized carriers into spin-based devices. Researchers turned
to doping small amounts (limited to a few percent) of transition metals into conventional
semiconductor materials like Si or GaAs, in an attempt to form what has now become
referred to as a dilute magnetic semiconductor (DMS) [7,8,12,13]. Figure 1.1 shows a
schematic of DMS material. Here, the unpaired d electrons on the transition metal dopant
ions would give rise to a magnetic moment. If these moments could be coupled
ferromagnetically then a magnetically ordered DMS systems would be available that
could be directly integrated with spin-based devices (e.g. LEDs and FETs) fabricated
from the same host semiconductor. This was expected to greatly reduce the scattering of
3
spins at the device interface that resulted from the presence of strain, density of state
discontinuities and lattice mismatch.
Figure 1.1: Schematic showing three types of materials a)Magnetic, b) Diluted
magnetic semiconductor and c) non magnetic semiconductor materials.[13]
Mn-doped II-VI compounds were the most studied DMS systems in the 1970s and
80s[14,15]. A tremendous amount of research was done on bulk II-VI semiconductors
like CdTe,CdSe, ZnTe, etc which could be substitutionally doped with Mn to vary the
electronic band structure and lattice constant. Interesting optical and electrical properties
including giant Faraday rotation, magnetic field induced metal-insulator transition and
formation of “bound magnetic polarons” were observed in these systems due to Zeeman
splitting of levels resulting from sp-d exchange interactions between the sp band and the
d electrons of Mn2+. However use of these materials in practical applications was limited
owing to difficulty of doping both p-type and n-type, and presence of predominantly
antiferromagnetic interactions between Mn spins. The II-VI semiconductors were favored
over III-V semiconductors because of their high solid solubility of transition metal atoms.
But, the advent of non-equilibrium crystal growth techniques like low-temperature MBE
and PLD allowed for the incorporation of magnetic dopants well beyond their
equilibrium limits, expanding the potential selection of semiconductor systems.
4
The interest shifted from bulk to thin film DMS materials as any device application
would require high quality epitaxial thin films. For the first time DMS epitaxial thin films
were grown of Mn doped InAs by low temperature MBE [16]. Although, homogenous
InMnAs films showed paramagnetic behavior, this work made an important contribution
by demonstrating the ability to grow good quality thin films with a high magnetic dopant
concentration. GaAs was also proposed to be potential candidate for DMS. This was of
particular interest since GaAs already had wide applications including, semiconductor
lasers, compact discs, microwaves, etc. In 1998 ferromagnetism above 100 K was
reported in GaMnAs [13]. This led to a flurry of studies, aimed at attaining room
temperature ferromagnetism in these systems. But, at this time the highest Curie
temperature(Tc) achieved in this system has been 172 K [17]. The low Curie temperature
hinders its use as practical material system for use in spintronic applications. More
recently, wideband gap oxides and nitrides have been studied extensively for use as DMS
materials. Transition metal doped GaN and ZnO thin films have generated special
interest since high temperature ferromagnetism with Tc > room temperature has already
been reported in these systems [18,19]. ZnO has aroused the most interest due to its
attractive optical, electrical and piezoelectric properties [20]. The high solubility of the
various 3d transition elements in ZnO adds further to its advantage over the III-V
semiconductors as higher concentrations of magnetic dopants can be introduced without
precipitating as secondary phases or metallic nanoclusters [21, 22].
Although room temperature ferromagnetism has been achieved in transition metal doped
ZnO, there exist a number of contradictory reports about the magnetic behavior of these
5
films. This mainly stems from the effect of growth conditions and technique on the
ferromagnetic ordering of spins. Therefore, it was realized that in order to achieve room
temperature ferromagnetism in ZnO based DMS, it was also important to understand the
primary mechanism governing the ferromagnetic ordering in these systems. A proper
understanding of the origin of any phenomenon is essential in order to achieve desired
properties for use in device applications. The origin of ferromagnetic ordering in ZnO is
still under debate. Both the carrier(electron) mediated exchange mechanism [19] and a
defect mediated mechanism like Bound Magnetic Polaron model(BMP) [23] have been
put forth as possible mechanisms responsible for the magnetic ordering. However, the
presence of magnetic secondary phases or nanoclusters has also been forwarded to
explain the observation of ferromagnetism in the DMS samples [24].
Finally, the main goal of developing a DMS materials system showing ferromagnetism
above room temperature is to directly incorporate them into spintronic based devices. The
main requirement for making such devices is efficient injection, transport,detection and
manipulation of spin-polarized carriers in semiconductors at room temperature [5].
In this dissertation, the growth and properties of Cu-doped ZnO based DMS films are
studied in great detail. The reason behind selecting Cu as a magnetic dopant are two fold
: a) Cu when existing in +2 state will have an electronic configuration of 3d9 . Hence, it
will have one unpaired electron and thus can potentially provide localized 3d spins for
long range ferromagnetic ordering. b) The fact that neither metallic Cu, nor its oxides
(CuO or Cu2O) are ferromagnetic [25], which rules out the possibility of ferromagnetic
signal generating in this system due to the presence of magnetic secondary phases or
6
nanoclusters. The effect of co-doping with different transition metals into ZnO has also
been studied by co-doping ZnO with Cu and Co. This exercise was not only useful in
determining the effect of co-doping on the ferromagnetic coupling between respective
dopants, but also provides an indication about the relative strengths of magnetic coupling
in different dopants. Of special scientific interest has been the determination of the origin
of the ferromagnetic ordering of the localized 3d spins. Additionally, Al was doped into
Cu:ZnO thin films to introduce free carriers (electrons). Detailed characterization and
property measurement of this system not only led to an improvement of the electrical
properties (essential for efficient spin-injection) but also provided great insight on the
role played by free carriers in mediating the ferromagnetic ordering. Thorough annealing
experiments in oxygen ambient were done followed by detailed characterization to
understand the role of native defects (particularly oxygen vacancies) in stabilizing the
ferromagnetic ordering. Finally, to determine the applicability of the Cu-doped ZnO
DMS system in spintronic applications, fabrication of spin-valve type device structures to
characterize the injection of spin polarized electrons into semiconductors using this
material system has been explored. The following chapters discuss in detail these
experimental results and their implication on the status of the field of DMS for spintronic11)

2 Chapter 2: General introduction to diluted magnetic
semiconductors
The earliest reports in the field of DMS date back to the 1960s. Extensive research was
done on Eu- and Cr-based chalcogenides (e.g EuSe, EuS, EuO, CdCr2Se4, and CdCr2S4).
[1- 5]. However, these systems proved to be very difficult to grow and possessed very
low Curie temperatures, thereby limiting their practical applications. Studies on the broad
class of II-VI compounds continued into the 1970s and 80s [6-12]. It was found that
materials systems like CdTe, CdSe, HgTe, ZnSe, ZnTe, etc could be substitutionally
doped with Mn to vary the electronic band structure and lattice constant, both
characteristics of substitutional Mn atoms in a AIIBVI lattice. Highly efficient
electroluminescence was achieved making them attractive for flat panel display
applications [13]. In addition, the Zeeman splitting of electronic levels that resulted from
the sp-d exchange interactions between the sp band and the d elelctrons of Mn2+ led to
interesting optical and electrical behavior including giant Faraday rotation, magnetic field
induced metal-insulator transition and formation of “bound magnetic polarons”.
However, except for the development of optical isolators the studies found few
applications because the magnetic interactions between Mn spins in these DMS systems
were predominantly antiferromagnetic, which resulted in paramagnetic,
antiferromagnetic or spin glass behavior[14]. Difficulty in doping these II-VI DMS to
obtain both p-type or n-type behavior was also experienced [15]. It should be noted that
most of this research was conducted on bulk materials. The II-VI semiconductors were
10
favored because of their high solid solubility of transition metal atoms. In comparison the
equilibrium solid solubility of these dopants in the III-V semiconductors is very low,
hindering the incorporation of higher concentrations of magnetic dopants. However, the
onset of non-equilibrium crystal growth techniques like low temperature MBE and PLD
allowed for the incorporation of magnetic dopants well beyond their equilibrium limits,
expanding the potential selection of semiconductor systems.
The first report of a thin film DMS was by Munekata et al. in 1989[16]. They studied Mn
doped InAs films grown by low temperature MBE in the temperature range of 200-300
0C. These films had n-type conductivity. Films grown at 200 0C were homogenous and
were found to be paramagnetic, while films grown at 300 0C showed ferromagnetic
behavior. However, the ferromagnetism was attributed to the presence of MnAs
nanoclusters. Despite of the fact that the homogenous InMnAs films showed
paramagnetic behavior, this work made an important contribution by demonstrating the
possibility of growing good quality thin films containing high magnetic dopant
concentrations. In 1992 Ohno et al. conducted magneto-transport studies on p-type
InMnAs thin films epitaxially grown by low temperature MBE [17]. This was the first
time that intrinsic ferromagnetism was observed in a III-V based DMS, albeit the
ferromagnetic ordering existed only below 7.5 K. In 1996 Ohno et al.[18] reported for the
first time the growth of GaMnAs DMS thin films. This was of particular interest since
GaAs was a well studied system that was widely used for optoelectronic applications.
Thus, ferromagnetism in GaAs would be very useful in making a functional material
system for spintronic applications. The Curie temperature was found to be ~ 60 K for a
11
Mn concentration of 5%. Then in 1998 Ohno reported ferromagnetism above 100 K in
p-type GaMnAs[19]. This led to a surge of studies attempting to push the Curie
temperature up to above room temperature. Unfortunately, to date, the highest reported
value of Tc for Mn doped GaAs is 172 K by Nazmul et al. [20]. Despite the advantages of
having good control over the growth and the magnetic properties of the III-As system, the
low Curie temperature hinders its use as practical material system for use in spintronic
applications. Thus, the search for other host semiconductor systems that would have
ferromagnetism at temperatures above room temperature continued.
A major advance in this search was achieved with the theoretical work of Dietl et al. [21].
They used the Zener model [22-24] of ferromagnetism to predict Curie temperatures for
TM doping of some widely used wide band gap semiconductors. Figure 2.1 shows
calculated Tc values for different semiconductors. The most promising are found to be
GaN and ZnO doped with 5% Mn, which showed Tc above 300 K. A critical requirement
assumed by the theory was the need for a very high hole concentration (3.5x1020 cm-3).
Dietl predicted that hole mediation was necessary to achieve the long range
ferromagnetic exchange between the dopant ion spins that was needed to achieve room
temperature ferromagnetism. This work stimulated a series of experimental research
efforts in search of room temperature ferromagnetism in wide band gap semiconductors,
including extensive studies on GaN and ZnO. Subsequently ferromagnetic ordering
above room temperature has been achieved in many of these systems.
12
Figure 2.1: Predicted Curie Temperatures of Mn-doped Wide Band gap
Semiconductors [21]
The first observation of room temperature ferromagnetism was reported by Reed et
al.[25] in GaMnN thin films grown by metalorganic chemical vapor deposition
(MOCVD) on (0001) sapphire substrates. Soon thereafter, Ueda et al.[26] examined the
n-type Zn1-xMxO(x=0.05-0.25) system (M= Co, Mn, Cr and Ni) grown by pulsed laser
deposition (PLD), and reported that only the Co-doped films displayed ferromagnetism
above room temperature. They found that very high carrier (electron) concentrations of
1020 cm-3 were required. However, the reproducibility of these films was reported to be
very poor (only 10%). Subsequent ab-initio studies predicted that n-type ZnO doped with
most transition metals (Co, Fe, Cr and Ni), should be ferromagnetic, but predicted that
holes would be required to mediate ferromagnetic exchange in Mn-doped ZnO[27,28].
The presence of room temperature ferromagnetism in ZnO (both predicted and
experimental reports) clearly extended the search beyond Dietl’s original Zener mean
13
field model prediction, and suggested that a high concentration of holes was an absolute
requirement for mediating ferromagnetic order above room temperature in the wide
bandgap systems. Dietl’s model [21] explained the onset of ferromagnetism in the Mn
doped III-As systems extremely well. Here Mn acts as an acceptor and the
ferromagnetism is driven by an exchange interaction between valence band holes and
localized spins of Mn ions. But in case of wide band gap semiconductors like GaN and
ZnO where p-type doping of the order mentioned in Dietl’s prediction (1020 cm-3) is
difficult to achieve, it is clear that strong ferromagnetic ordering cannot be due to
exchange interaction with holes. In fact, this appears to be the case for most of the reports
on ferromagnetism in wide band gap oxide systems. ZnO, TiO2 and SnO2-based DMS are
all n-type materials [19,29,30]. Initial studies suggested a free carrier mediated
mechanism to be responsible for ferromagnetic order in DMS [26-28]. But some recent
reports have observed ferromagnetism in systems that have relatively low carrier
concentrations [31,32]. Here it appears that intrinsic defects like oxygen vacancies are
playing a very important role in stabilizing ferromagnetic order in these systems [33-35].
In some cases, the presence of secondary phases or nanoclusters of magnetic phases in
the host system could also explain the observed ferromagnetic signal [36]. But, in either
case, it remains to be determined whether there exists any degree of coupling between the
itinerant carriers and the localized magnetic spins. Such a coupling must exist if these are
to be used as an efficient spin injector i.e. the ferromagnetism needs to be strongly
polarizing the itinerant carriers if efficient spin injection is to be realized.
14
For a DMS material to be useful in spintronics applications it is necessary that the
ferromagnetic behavior should be an inherent property of the whole volume of the DMS
and not due to the presence of ferromagnetic nanoclusters irrespective of whether they
occur due to the presence of impurities during growth, precipitation of magnetic
secondary phases or nanoclusters or spinodal decomposition leading to formation of nano
regions rich in the magnetic dopant. Even though they might show high temperature
ferromagnetic ordering, this would have very weak influence on the bulk properties of the
DMS such as magnetoresistance, magnetooptical characteristics and anomalous Hall
effect. It is also necessary that the ferromagnetic ordering in these DMSs should exist at
temperature much higher than 300 K so that they can be used for injection of spin
polarized carriers at practical temperatures. The fabrication technology of these
ferromagnetic semiconductors should be robust and the characteristics should be
reproducible to a large extent. This is necessary because the literature is fraught with
reports showing different magnetic characteristics for the same material system that
appear to depend on the growth condition. For a DMS to be used in large scale
commercial production of spintronic devices, they should not show large variation in
behavior with small changes in fabrication conditions. Also necessary is the ability to
grow these DMS thin films epitaxially on common substrates so that they can be easily
integrated with commonly used semiconductors in microelectronic devices such as Si,
GaAs,etc. Based on these requirements the challenges that are faced by existing
ferromagnetic DMSs can be summarized as follows [37]:
15
i. Low solubility of transition metals in the semiconductor host – The excess
concentration may lead to formation of magnetic nanoclusters which are
responsible for the ferromagnetic signal in these materials
ii. Low temperature of Ferromagnetic ordering - For practical applications it is very
important that the long range ferromagnetic ordering in these systems be retained
at temperatures higher than room temperature.
iii. Reproducibility of results - For commercial applications, it is very important that
we have high a degree of reproducibility of properties of the materials and that
they should not vary drastically with small changes in growth parameters. To
have a better control over the magnetic properties of DMS material systems, it is
necessary that there should be a clear understanding of the origin of
ferromagnetism in these material systems. There still exists a debate over the
mechanism governing long range ferromagnetic order in the DMSs
iv. Fabrication of high quality thin films - To be integrated with currently existing
semiconductor technology it is essential that high quality single crystal thin films
of the DMSs can be grown epitaxially on common substrates such as sapphire and
silicon.
In the following sections we will briefly discuss about the transition metal doped ZnO
based DMS systems. Special emphasis will be on Mn, Co and Cu–doped ZnO.
Subsequently a few mechanisms that are operative in these DMS material systems will be
reviewed. And finally, some potential device applications will also be discussed.
16
وربي يعافيكم ويخليكم قولوا امــــــــــــــــــــــــــــــــــــــــــــــــ ــــــــين