The high crystallinity, highly ordered, exceptional morphology, and

The magnetic, optical and electronic properties of nanoparticles are dependent on their size, shape, structure and composition. The adjustable size and high surface area to volume ratio allow tailoring of these properties in highly fashioned order and thus can be tuned for the desired application. This flexibility of nanoparticles has attracted potential applications in fields like, biomedical, nano-electromangetic, photonic, plasmonic, electronic, and spintronics. The nanoparticles are used in data storage devices and spin transport devices for their small size and single domain magnetic properties. They are also used in drug delivery, magnetic resonance imaging (MRI) for their low toxicity, as high contrast agent and are best suited for biomedical applications. The nanoparticles are fundamental parameter and building blocks of new materials and behave like an additional aspect in the periodic table. Magnetic nanoparticles are good candidate for the study of magnetic properties for various applications including biomedical, enhanced oil recovery, magnetic energy materials, nano-electronics and spintronics, high-density data devices. Magnetic nanoparticles show remarkable magnetic properties such as superparamagnetism, magnetic hyperthermia and tunable relaxation time. These excellent properties make them ideal to accomplish fundamental studies when compared with their bulk counterparts. One of the key considerations is conversion of magnetic energy to the thermal energy. This heat transfer of magnetic nanoparticles is being used as a capable mechanism for many applications. The physical properties are dependent on the size of an object and varies along with relevant property, for example in case of magnetism, it depends on the exchange length or the size of the magnetic domains (10-1000nm). Such characteristics makes the magnetic nanoparticles unique from their bulk counterparts and their physical properties build whole new field of magnetism. Reducing the size of magnetic nanoparticles, domains stabilize their energy from multidomain to single domain and magnetic properties move, for example, from ferromangetism to superparamagnetism. The critical radius of magnetite is 60 nm, approximately where the magnetic properties transfer from one to another regime. The spin in the single domain nanoparticles is often called as super-spin or super-moment or giant-moment.Such single domain nanoparticles or giant-spins are present in living organisms like bacteria and ferritin. They produce nanoparticles of distinct size, high crystallinity, highly ordered, exceptional morphology, and large monodispersity through biomineralization process. Biomineralization process involves the chemistry of inorganic matter with the organic substances for various purposes including the iron storage, structural support, and magnetic interactions. The resulting products are highly stable in harsh environments like high and low temperature, pressure and pH values. These naturally occurring biominerals, like iron oxide nanoparticles, outweighs the laboratory synthesized nanoparticles due to its characteristics. The best known magnetic biominerals are the iron oxide nanoparticles ($Fe_3O_4$) in magnetotactic bacteria. Efforts are being put in biomimicry and bio-inspired synthesis of nanoparticles and their assembly in nanostructures and superlattices.The membrane enclosed magnetic nanoparticles inside magnetotactic bacteria are known as magnetosomes, mostly aligned in a chain form, constructing a magnetic compass (a bar magnet) for the orientation along the external magnetic field (naturally earth’s magnetic field). This allignment of microaerobic bacterium along the earth’s magnetic field is called as magnetotaxis, discovered by R. Blakemore in 1975. The magnetosomes crystallize themselves in <111> direction for enhanced mangetic moment per particle, and these magnetic moments of individual particles align along the chain direction and build a giant magnetic moment. The resulting torque on the chain, due to giant magnetic moment and external magnetic field, redirects the bacterium to change its direction of motion. The chain like structure of magnetosomes inspired many researchers to artificially assemble the nanoparticles in 1D, 2D and 3D arrays. The magnetosomes by themselves have also been assembled in 2D and 3D arrays on millimeter and micrometer scales, using external electromagnets. Nevertheless, there are challenges still to be explored for well-defined positioning, orientation and assembly on nanoscale dimensions. It requires modern nanofabrication techniques to be used for control experiments and for governing the geometry shape and interparticle distance. The magnetism of random assembly of magnetic nanoparticles depends on the type in strength of dipolar coupling and interactions. The organized assembly would involve the geometrical anisotropy parameters along with the dipolar strength. The magnetism of supersymmetrical shapes of magnetic nanoparticle assemblies is also often called supermagnetism. The correct positioning of nanoparticles at defined locations on the substate is essential  for further development. Various methods has been adopted in this regard during past few years, including the top down and bottom up approaches. The self assembly could be direct or indirect, force mediated, lithographically controlled, or by selective surface wetting. For uncharged, and non-magnetic nanoparticles, the intermolecular interactions and van der Waals forces drive the self assembly. For charged and magnetic nanoparticles, the electrostatic and magnetic forces also take part in the self assembly along with the fluid dynamical forces. The direct self assemblies include the micro-contact printing and scanning probe-based lithographies. The assisted self assemblies demonstrate the pattering on templated surfaces using capillary forces, and convective flows. The templates could be made using optical lithography, electron beam lithography or the ion beam lithography. The template assisted self assembly could be wet or dry; the wet method involves the capillary forces of the nanoparticle suspension; the dry method uses the van der Waals forces into the grooves of the template. The template assisted assembly is most commonly used because of its low cost and high throughput. Using these approaches, many forms of assemblies have been achieved to date on the nanometer scale. The 2D suplattices and long-range orders can be developed by the drop-cast on the substrate surface, where long-range order depends on the evaporation rate of the fluid. Aqueous interfaces are reported for the long range self assembly of nanoparticles that can be conveyed to the substrate surface. The resulting arrays are of exceptional with unique electronic and magnetic properties. These 2D arrays are of immense interest because of their unique properties. The study of 2D arrays of magnetic nanoparticles has been studied theoretically but it has to be explored experimentally. The collective magnetic properties of 2D arrays of magnetic nanoparticles are different from their bulk counterparts or the individual nanoparticles in theoretical predicitons. These arrays could accomplish the new properties not found the bulk material. But the assemblies the magnetic nanoparticles on the nanomechanical devices have never been reported. Patterning 2D arrays of nanoparticles in geometrical shapes on top of the nanomechanical and optomechanical devices is aimed in this Thesis. The number of nanoparticles and the geometrical arrangement of array can be controlled, stabilized and it gives reproducibility and high throughput that could be extend to the wafer-scale mass production of devices. subsection{Torque Magnetometry}A magnetometer can measure the magnitude, direction or the relative change of magnetization, or the magnetic field of a material. Beside the magnetic compass, the earliest magnetometer was invented in 1833 by Carl Friedrich Gauss for the earth’s magnetic field measurement. Measuring the magnetization by the magnetometer gives us the information about the magnetic ordering and characteristics of the magnetic material. The efficiencies and capabilities of magnetometers are dependent on the sensitivities, noise detection, drift, stability, resolution, sampling rate and bandwidth. The applicable magnetometers are SQUIDs, vibrating sample magnetometer (VSM), inductive coils magnetometers, Faraday force magnetometry, optical Magnetometry (magneto-optic Kerr effect; MOKE), Hall effect magnetometer, magnetoresistive devices, fluxgate magnetometer and torque magnetometers. SQUIDs (superconducting quantum interference devices) are vector magnetometers, sensitive up to $sim 0.4 ext{ fTHz}^{-1/2}$ noise level but requires low temperature for the measurements for cooling the superconductors. Inductive coil magnetometers works on the induction principle and are only limited to the ac magnetization detection. In vibrating sample magnetometers (VSM), the sample vibrates inside an inductive coil and change in flux is measured. It is limited by the sample vibration and heat, not useful for delicate samples, and an order of magnitude less sensitivity than the SQUIDs. Faraday force magnetometry requires the magnetic field gradients in the measurements and it is also an order of magnitude less sensitivity than the SQUIDs. Magneto-optic Kerr effect (MOKE) magnetometry and the Faraday rotation magnetometry use optical polarization for magnetization measurements. Magnetometers can also be used for field sensing, for example most common field sensors are, Hall effect sensors, magnetoresistive sensors, and fluxgate sensors. Magnetometers are being employed in various field, for example, in  archaeology, auroras, geology in geophysics, military, mobile phones and electronics, spacecraft and many other fields. Beside these methods, there are microscopic methods for measuring the magnetization. Inspired from atomic force microscopy (AFM), there is magnetic force microscopy (MFM) where magnetic cantilevers sense the magnetic fields. The cantilever motion can be static or dynamic (AC driven) for the magnetic force detection. The MFM technique was extended to the high resolution magnetic resonance force microscopy (MRFM) to electron and nuclear magnetic resonance imaging and detection, capable of single electron spin detection and extra-ordinary spatial resoltuion (10nm). Another extension of these methods is the ferromagnetic resonance force microscopy (fMRFM) used for spin wave modes and magnetization dynamics studies. Magnetic torque magnetometry is invasive technique with outstanding sensitivity, for measuring torque that is proportional to the magnetization of the magnetic material. The magnetic torque is produced by the equation, $ au = mu imes H$, where H is external magnetic field. The resulting torque is the measure of net magnetization of the sample. However, torque magnetometry is cable of resolving and measuring the magnetization components in vector form, so that the orientation bound properties of the material can be extracted. The magnetic torque magnetometry is a wide platform that can be used from spin resonances to the quantum oscillations or phase transition studies. The actuation of magnetic processes could be on the cantilever, and read-out could be performed by the electrical capacitance, optical interferometry, or the laser taper fiber methods in the cavity optomechanical systems. Torque magnetometry is not in direct competition with existing methods, but offers a complementary magnetometry tool at the nanoscale. In comparison to most other magnetometry methods involving nanodevices our torque magnetometry method provides direct, non-invasive, and fast acquisition of the magnetostatic hysteresis loop while also being able to capture the associated RF susceptibility. Magnetic force cite{Mamin2009} and diamond NV centre cite{Degen2008, Taylor2008, Maze2008, Balasubramanian2008, Maletinsky2012, Rondin2013, Tetienne2014, Staudacher2013, Grinolds2013, Muller2014, Rugar2015, Pelliccione2016} magnetometry offer extremely high magnetic moment sensitivity for electron and nuclear spin resonance detection, and they are practically ideal for localized probing. However, experimental acquisition of the volumetric static moment of a micromagnetic element (and acquiring its hysteresis loop) would require lengthy imaging and reconstruction. This is further complicated if complex three-dimensional microstructures are to be measured (of which torque magnetometry is capable cite{Losby2015}).Micro-SQUID has also recently achieved single spin sensitivity cite{Vasyukov2013}, though their limited operating temperature does not allow for room-temperature measurement. Planar micro-Hall measurements that are sensitive to the perpendicular component of stray field offer room temperature sensitivities near that of nanoscale torque magnetometers, and have measured Barkhausen signatures associated with vortex core pinning in fabricated defect sites cite{Rahm2003}. However, Johnson noise dominates and limits their detection sensitivity at high frequencies, and there have been no reports, to the knowledge of the authors, of RF susceptibility measurements associated with Barkhausen signatures in single nanoscale elements using the micro-Hall method. Recently, inductive methods for the sensitive measurement of magnetic resonance in single nanoscale elements have been developed cite{Banholzer2011,Tamaru2014}. For inductive measurement of the irreversible magnetization changes at the static limit (DC), superconducting electronics would be required cite{Richardson1988}.The measurement capabilities are improved by further miniaturizing the device dimensions or using the cavity optomechanical platform for higher frequencies and higher sensitivities. Furthermore, the AC lock-in technique could be combined for enhanced sensitivity at the natural resonance frequency of the device. The toque magnetometers are employed in this Thesis, using the nanomehcanical and optomechanical platforms. The concepts in this subsection and described in the above subsection would be combined in the next section.