«c SIMON LINDBERG, 2013 Cover: To the left is an image of a mono crystalline silicon solar cell , In the middle is an image of three samples of ...»
2.1 Overview Metal hydrides have in general been of most interest as potential candidates for hydrogen storage but quite lately researchers has started to look into other properties such as electrical properties and optical properties. Mg2 NiH4 is no exception and it has been extensively investigated for its hydrogen storing properties mostly in the form of nano particles, but the theory of semiconducting metal hydrides were ﬁrst presented in 1984 by Mich`le Gupta . The ﬁrst experimental e work on semiconducting thin ﬁlms was presented in 2001 by Richardson et al at Lawrence Berkeley National Laboratory. The group were primarily researching if Mg2 Ni could be used as a switchable mirror in the presence of hydrogen and they discovered that the ﬁlm switched from a shiny metallic state to a transparent semiconducting state in the presence of hydrogen.
The electrical properties of palladium capped Mg2 NiH4 have investigated by Enache et al in 2004 at the Vrije University , but no research of electrical properties has been done on uncapped ﬁlms.
In 2008 a paper regarding the calculated electronic structure of Mg2 NiH4 was published by Karazhanov and Ulyashin where they compared theoretical values of Mg2 NiH4 and silicon which showed promising similarities between Mg2 NiH4 and silicon.  In 2011 Mongstad et al showed that it was possible to synthesize thin ﬁlms of Mg2.x NiH4 using reactive sputtering for the ﬁrst time, previously most methods had involved a palladium capping layer. In 2012 the band gap and optical properties were determined experimentally by J.H. Selj et al which conﬁrmed the previous calculations done on Mg2.x NiH4 and concluded that the band gap varies between
1.6eV-2.1eV for different structures of Mg2.x NiH4. Research is still lacking in the areas of semiconducting properties such as carrier types and mobility for uncapped thinﬁlms.
The structures of Mg2 NiH4 has been quite extensively researched previously, related to studies of the hydrogen storage properties of the material but not in the form of thin ﬁlms. The vibrational spectra of Mg2 NiH4 has also been investigated in the form of nano particles , but any research of the thin ﬁlm Raman spectra is still missing.
2.2 Introduction to the novel semiconductor Mg2NiH4 An exciting new material which has the prospect of fulﬁlling some of the criteria mentioned previously is Mg2.x NiH4, which combines the positive properties of traditional silicon solar cells such as relatively low material cost with the positive properties of thin ﬁlm solar cells which are high absorption and low material usage. There are also other advantages such as a very appropriate band gap of 1.6eV for certain structural conﬁgurations which are quite close to the optimal band gap of 1.4eV for photovoltaic applications, simpler and cheaper production methods compared to both silicon and thin ﬁlm solar cells.
As mentioned before Mg2 NiH4 can obtain three different structures which affects the electrical properties of the material. The ﬁrst structure is the amorphous as-deposited structure of sputtered thin ﬁlms. This amorphous structure lacks long ranging periodicity which makes structural characterization difﬁcult but TEM measurements done in previous papers shows that it is actually nano crystalline which means that the materials are consistent of small crystallites with uniform structure but the crystallites are arranged independently to each other which is why there is no long ranging symmetry. The second is the so-called low-temperature (LT) phase it forms when synthesizing Mg2 NiH4 -powders below 510K. This LT phase has a monoclinic structure, where the magnesium atoms are located in the corners and the NiH4 complex is in the middle. The third structure is the high temperature phase which is achieved when heating up the powder to 510K, it retains the same location of the atoms as in the LT-phase, but changes from a monoclinic to a cubic structure.
Figure 2.1: The cubic FCC structure of Mg2 NiH4 .
Purple balls represent Mg, green balls represent Ni and gray are the average positions of hydrogen.
The HT phase is found in Figure 2.1, in Figure 2.1 are only the average positions of the hydrogen atoms marked, which have been established by neutron scattering experiments on Mg2 NiD4 in room temperature . Calculations also show that the hydrogen should form distorted tetrahedral structures around the nickel atoms, similar to the structure seen in Figure 2.2.
The difference between the cubic and the monoclinic phase is demonstrated in Figure 2.2, in this ﬁgure the hydrogen atoms also assumes more likely positions compared to the structure in Figure 2.1. In thin ﬁlms the structural behavior is different from nanoparticles, as mentioned before the as-deposited structure is nanocrystalline and when heated up to 590K it undergoes an irreversible change to the cubic HT-phase if there is an excess of magnesium in the samples, the LT-phase has not been detected in thin ﬁlms yet.
Figure 2.2: Schematic image that demonstrates the difference between cubic (a) and monoclinic (b) structures.
The physical properties that govern the semiconducting properties of Mg2 NiH4 was ﬁrst explained by Gupta et al , but is also discussed by Garcia et al , Karazhanov et al  and by Enache et al  among others. The general concept of the theoretical work states that the magnesium atoms form Mg-Mg bonds in the material which are thought to form the conducting part in the semiconductor while the NiH4 -complexes will affect the orbitals which opens up the band gap, which means that the material becomes semiconducting. The papers above are focused on the crystalline structures and do not address the amorphous structure.
2.3 Synthesis and sample preparation 2.3.1 Reactive sputtering The thin ﬁlms were deposited using a reactive sputtering device by Leybold sputter system A550v7. The sputtering takes place in a vacuum chamber which is ﬁlled by a gas mixture of argon and hydrogen up to a pressure of 2·10−2 mbar. A strong electrical ﬁeld is then established between the targets and the substrate which creates a plasma of Ar+ - and H+ ions, where the positively charged Ar-ions are attracted to the negative surface of the metal targets and hits the surface. The kinetic energy from the Ar-ions are transfered to the target atoms and are kicked out onto the substrate.
For synthesis of the ﬁlms co-sputtering with two targets activated simultaneously was used to create the desired compositions; a magnesium-target and a Mg2 Ni-target. For co-sputtering in the system used there were two different power sources: A DC-source and a RF-source, so for the depositions in this work the Mg-target was powered by the RF-source and the Mg2 Ni-target was powered by the DC-source. The reason why this setup was chosen was because Ni is magnetic and thus can be effected by the oscillating ﬁeld, while magnesium is not.
The carrier was set to oscillate in front of the targets with a speed of 0.2 m/s, a schematic of the setup is shown in Figure 2.3. The general process parameters were set according to previous work done with metal hydrides  and the power of the magnetrons were adjusted to achieve the different magnesium ratios.
Figure 2.3: A schematic image of the processing chamber.
For EDS, XRD and Raman measurement a layer of Mg2.x NiH4 were deposited on clean glass surfaces and then cut to the appropriate geometry. For the Hall and resistivity measurements Figure 2.4: Schematic overview of the samples for electrical characterisation.
contacts were also necessary, these were made out of aluminum and the processing parameters for the contacts can also be found in Table 3.1, a schematic of the sample can be found in Figure 2.4. In this case the samples were ﬁrst cut out, then cleaned before deposition of the semiconductor ﬁlm and ﬁnally contacts were deposited using a mask made out of single crystal silicon wafers. Two different geomtries were used, the simplest was the van der Pauw-geometry which is seen in Figure 2.4. The second was a so called bridge type geometry which is shown in Figure 2.5. This sample was created using a template made out of an alumina wafer in which the bridge shape was cut with a laser.
Figure 2.5: Schematic overview of the bridge type-sample used in the PPMS.
2.4 Characterization 2.4.1 Raman spectroscopy In the case of Mg2 NiH4 hydrogen plays an important role, so to determine where it is situated in the material is very important. However hydrogen is very difﬁcult to detect with most methods, so Raman spectroscopy was used because it gives the opportunity to detect hydrogen indirectly by the bonds to the other atoms in the material.
Raman spectroscopy is a vibrational method based on the Raman process which is inelastic collisions between photons and molecules or atoms in a material, what happens is that the system absorbs the photon and is excited to a high, ’virtual’, energy level. The system then instantly relaxes back to a lower energy level at the same time it emits a photon. When the the system
relaxes back two different scenarios can occur:
1. It can relax back to the same energy it had before being excited and emit a photon which has the same energy as the incoming photon. In this case the collision is elastic and the radiation that is released is called Rayleigh scattering.
2. It is also possible that it relaxes back to an energy level with a different energy and emits a photon with changed energy compared to the incoming photon. In this case there are two options: if energy is transfered from the photon to the system the process is called Stoke’s scattering while if energy is transfered to the photon it’s called anti-Stoke’s scattering.
Of the options above it is the second which is interesting in the case of Raman spectroscopy, a graphical representation of the processes can be seen in Figure 2.6. From the energy difference measured in the scattered photons together with the knowledge of the structure, bonds and atomic species present in the material it is possible to assign the peaks in a Raman spectra to certain motions such as vibrations and translations and determine which atomic species that are involved.
Figure 2.6: The different interband excitation alternatives in atom-photon interactions.
 However, not only atomic excitations are possible to occur when light interacts with matter, light can also interacts with the electrons present in the sample. In a semiconductor this interaction between the photons and the electrons is governed by the relationship between the band gap and the laser used in the Raman instrument. If the energy of the laser is higher than the band gap then electrons will be excited to the conduction band, and this could change the appearance of the spectra. The second alternative is if the energy of the laser is smaller than the band gap no electrons will be excited, this is called Raman resonance mode and in this case more modes could be present in the spectra. On a ﬁnal note, it is also possible to detect other physical properties of the material such as plasmons, crystallite size and in some cases the band gap itself might be visible.
For more information regarding Raman spectroscopy and confocal Raman spectroscopy see the book ’Confocal Raman Microscopy’ by Thomas Dieing .
2.4.2 Hall-measurements Hall-measurements is the most commonly used technique to determine physical properties such as mobility, resistivity, charge-carrier density and majority charge carriers in semiconductors.
These can all be determined in one single measurement.
As mentioned before in Chapter 1.2 there are two types of charge-carriers in semiconductors and to be able to optimize a photovoltaic cell from a speciﬁc semiconductor material the majority charge carrier in that speciﬁc material needs to be determined.
The Hall-effect was discovered in 1879 by Edwin Herbert Hall, and is in short the effect a magnetic ﬁeld has on the current ﬂowing in a conductor or semiconductor. In a conductor without any inﬂuence of a magnetic ﬁeld the electrons ﬂow in a straight line from one end to the other.
While if a strong magnetic ﬁeld is introduced perpendicular to the current, the electrons will be inﬂuenced by the so called Lorentz force which will effectively alter the movement. This change can be described as if the electron trajectories are slightly bent, which is illustrated in Figure
2.7. This change also lead to an accumulation of charge carriers on one side of the material thus inducing an electrical ﬁeld perpendicular to the applied current and the magnetic ﬁeld, which is called the Hall-voltage. A schematic view of a Hall-setup can be found in Figure 2.7. If the Figure 2.7: Illustration of the general setup for Hall-measurements, I = The current sent through the sample, VX =Voltage applied through the sample, BZ = Magnetic ﬁeld perpendicular to the current, VH = The measured Hall-voltage, t = thickness of sample. 
Where n is the density of charge-carriers and ﬁnally Equation 2.2 shows that the Hall-coefﬁcient will shift sign depending on the sign of the charge-carrier. For electrons the charge, e, is positive by deﬁnition and for holes the charge becomes negative.
So by knowing the current fed to the sample and the magnetic ﬁeld while measuring the transverse electrical ﬁeld, Ey, is it possible to determine the type of carriers in the material by looking at the sign of the Hall-coefﬁcient and the concentration of charge carriers depending on the magnitude of the coefﬁcient.  Some samples though can be problematic, if the carrier concentration is high the coefﬁcient is very small and combined with a small current jx can effectively lower the Hall-coefﬁcient below the sensitivity range of the device.
For more background and theory regarding Hall-effect and charge carriers can be found in the book ’Semiconductor materials: an introduction to basic principles’ .
Two different setups were used to try and determine the Hall-effect and properties.