Process Development of Aluminum Electroplating from an Ionic Liquid on 150 mm Wafer Level.

This paper focuses on the development of electroplating on 150 mm wafer level for microsystem technology applications from 1-Ethyl-3-methylimidazolium chloride (EMImCl) with Aluminumtrichloride (AlCl3). The deposition was carried out on 150 mm wafers with Au or Al seed layers deposited by physical vapor deposition (PVD). The electrodeposition was carried out using pattern plating. On the Au seed layer, bipolar pulse plating was applied. Compared to the Au seed layer, the electrodeposition on the Al seed layer was favorable, with lower current densities and pulsing frequencies. Utilizing the recurrent galvanic pulses and avoiding ionic liquid convection, inhomogeneities lower than 15% were achieved with a laboratory plating cell. One major aspect of this study was the removal of the native Al oxide prior to deposition. It was investigated on the chip and wafer levels using either current- or potential-controlled removal pulses. This process step was affected by the plasma treatment of the wafer, thus the surface free energy, prior to plating. It turned out that a higher surface free energy hindered proper oxide removal at a potential of 3 V. The theory of oxide breakdown based on electrostriction force via the electrical field was applied to discuss the findings and to derive conclusions for future plating experiments.

Low temperature atomic layer deposition of cobalt using dicobalt hexacarbonyl-1-heptyne as precursor.

In this work, we present the development of an atomic layer deposition (ALD) process for metallic cobalt. The process operates at low temperatures using dicobalt hexacarbonyl-1-heptyne [Co2(CO)6HC≡CC5H11] and hydrogen plasma. For this precursor an ALD window in the temperature range between 50 and 110 °C was determined with a constant deposition rate of approximately 0.1 Å/cycle. The upper limit of the ALD window is defined by the onset of the decomposition of the precursor. In our case, decomposition occurs at temperatures of 125 °C and above, resulting in a film growth in chemical vapour deposition mode. The lower limit of the ALD window is around 35 °C, where the reduction of the precursor is incomplete. The saturation behaviour of the process was investigated. X-ray photoelectron spectroscopy measurements could show that the deposited cobalt is in the metallic state. The finally established process in ALD mode shows a homogeneous coating at the wafer level.

Millisecond flash lamp curing for porosity generation in thin films.

Flash lamp annealing (FLA) with millisecond pulse durations is reported as a novel curing method for pore precursor's degradation in thin films. A case study on the curing of dielectric thin films is presented. FLA-cured films are being investigated by means of positron annihilation spectroscopy (PAS) and Fourier-transform infrared (FTIR) spectroscopy in order to quantify the nm-scale porosity and post-treatment chemistry, respectively. Results from positron annihilation reveal the onset of the formation of porous voids inside the samples at 6 ms flash treatment time. Moreover, parameter's adjustment (flash duration and energy density) allows for identifying the optimum conditions of effective curing. Within such a systematic investigation, positron results indicate that FLA is able to decompose the porogen (pore precursors) and to generate interconnected (open porosity) or isolated pore networks with self-sealed pores in a controllable way. Furthermore, FTIR results demonstrate the structural evolution after FLA, that help for setting the optimal annealing conditions whereby only a residual amount of porogen remains and at the same time a well-densified matrix, and a hydrophobic porous structures are created. Raman spectroscopy suggests that the curing-induced self-sealing layer developed at the film surface is a graphene oxide-like layer, which could serve as the outside sealing of the pore network from intrusions.

Nanolithographic Fabrication Technologies for Network-Based Biocomputation Devices.

Network-based biocomputation (NBC) relies on accurate guiding of biological agents through nanofabricated channels produced by lithographic patterning techniques. Here, we report on the large-scale, wafer-level fabrication of optimized microfluidic channel networks (NBC networks) using electron-beam lithography as the central method. To confirm the functionality of these NBC networks, we solve an instance of a classical non-deterministic-polynomial-time complete ("NP-complete") problem, the subset-sum problem. The propagation of cytoskeletal filaments, e.g., molecular motor-propelled microtubules or actin filaments, relies on a combination of physical and chemical guiding along the channels of an NBC network. Therefore, the nanofabricated channels have to fulfill specific requirements with respect to the biochemical treatment as well as the geometrical confienement, with walls surrounding the floors where functional molecular motors attach. We show how the material stack used for the NBC network can be optimized so that the motor-proteins attach themselves in functional form only to the floor of the channels. Further optimizations in the nanolithographic fabrication processes greatly improve the smoothness of the channel walls and floors, while optimizations in motor-protein expression and purification improve the activity of the motor proteins, and therefore, the motility of the filaments. Together, these optimizations provide us with the opportunity to increase the reliability of our NBC devices. In the future, we expect that these nanolithographic fabrication technologies will enable production of large-scale NBC networks intended to solve substantially larger combinatorial problems that are currently outside the capabilities of conventional software-based solvers.

Modeling the temporal evolution and stability of thin evaporating films for wafer surface processing.

The interaction of thin evaporating fluid films with solids is studied using the example of water on LiTaO3 (LTO). Adsorption energies are computed by ab initio density functional theory (DFT) and used to calculate the Gibbs free energy of adsorption of water on LTO. Integrating the disjoining pressure, consisting of molecular and structural components, with respect to film thickness gives an expression for the Gibbs free energy. In this way, parameters for the disjoining pressure can be calculated by fitting its integral to the Gibbs free energy computed by ab initio DFT. A combination of literature-known models for spin drying and evaporation is utilized to describe the temporal evolution of the water layer. The vapor above the water layer is modeled by diffusion and a mass balance is applied at the water-air interface. For thick initial layers, an analytical approximation is derived which only depends on fluid and ambient conditions but not on the substrate properties.

Laser induced crystallization of Co-Fe-B films.

Local crystallization of ferromagnetic layers is crucial in the successful realization of miniaturized tunneling magnetoresistance (TMR) devices. In the case of Co-Fe-B TMR devices, used most successfully so far in applications and devices, Co-Fe-B layers are initially deposited in an amorphous state and annealed post-deposition to induce crystallization in Co-Fe, thereby increasing the device performance. In this work, first direct proof of locally triggered crystallization of 10 nm thick Co-Fe-B films by laser irradiation is provided by means of X-ray diffraction (XRD) using synchrotron radiation. A comparison with furnace annealing is performed for benchmarking purposes, covering different annealing parameters, including temperature and duration in the case of furnace annealing, as well as laser intensity and scanning speed for the laser annealing. Films of Co-Fe-B with different stoichiometry sandwiched between a Ru and a Ta or MgO layer were systematically assessed by XRD and SQUID magnetometry in order to elucidate the crystallization mechanisms. The transformation of Co-Fe-B films from amorphous to crystalline is revealed by the presence of pronounced CoFe(110) and/or CoFe(200) reflexes in the XRD θ-2θ scans, depending on the capping layer. For a certain window of parameters, comparable crystallization yields are obtained with furnace and laser annealing. Samples with an MgO capping layer required a slightly lower laser intensity to achieve equivalent Co-Fe crystallization yields, highlighting the potential of laser annealing to locally enhance the TMR ratio.

Wafer-level integration of self-aligned high aspect ratio silicon 3D structures using the MACE method with Au, Pd, Pt, Cu, and Ir.

The wafer-level integration of high aspect ratio silicon nanostructures is an essential part of the fabrication of nanodevices. Metal-assisted chemical etching (MACE) is a promising low-cost and high-volume technique for the generation of vertically aligned silicon nanowires. Noble metal nanoparticles were used to locally etch the silicon substrate. This work demonstrates a bottom-up self-assembly approach for noble metal nanoparticle formation and the subsequent silicon wet etching. The macroscopic wafer patterning has been done by using a poly(methyl methacrylate) masking layer. Different metals (Au, Pt, Pd, Cu, and Ir) were investigated to derive a set of technologies as platform for specific applications. Especially, the shape of the 3D structures and the resulting reflectance have been investigated. The Si nanostructures fabricated using Au nanoparticles show a perfect light absorption with a reflectance below 0.3%. The demonstrated technology can be integrated into common fabrication processes for microelectromechanical systems.

Tunable Magnetic Vortex Dynamics in Ion-Implanted Permalloy Disks.

Nanoscale, low-phase-noise, tunable transmitter-receiver links are key for enabling the progress of wireless communication. We demonstrate that vortex-based spin-torque nano-oscillators, which are intrinsically low-noise devices because of their topologically protected magnetic structure, can achieve frequency tunability when submitted to local ion implantation. In the experiments presented here, the gyrotropic mode is excited with spin-polarized alternating currents and anisotropic magnetoresistance measurements yield discrete frequencies from a single device. Indeed, chromium-implanted regions of permalloy disks exhibit different saturation magnetization than neighboring, non-irradiated areas, and thus different resonance frequency, corresponding to the specific area where the core is gyrating. Our study proves that such devices can be fabricated without the need for further lithographical steps, suggesting ion irradiation can be a viable and cost-effective fabrication method for densely packed networks of oscillators.

Gate Spacer Investigation for Improving the Speed of High-Frequency Carbon Nanotube-Based Field-Effect Transistors.

Carbon nanotube (CNT)-based field-effect transistors have demonstrated great potential for high-frequency (HF) analog transceiver electronics. Despite significant advancements, one of the remaining challenges is the optimization of the device architecture for obtaining the highest possible speed and linearity. While most studies so far have concentrated on symmetrical top gated FET devices, we report on the impact of the device architecture on their HF performance. Based on a wafer-level nanotechnology platform and device simulations, transistors with a buried gate having different widths and positions in the FET channel have been fabricated. Analysis of several FETs with nonsymmetrical gate electrode location in the channel revealed a speed increase of up to 18% measured by the external transit frequency fT and maximum frequency of oscillation fmax. Although only randomly oriented CNTs with a density of 25 CNTs/µm and 280 nm long channels were used in this study, transit frequencies up to 14 GHz were obtained.

Charged domains in ferroelectric, polycrystalline yttrium manganite thin films resolved with scanning electron microscopy.

We have investigated ferroelectric charged domains in polycrystalline hexagonal yttrium manganite thin films (Y1Mn1O3, Y0.95Mn1.05O3, Y1Mn0.99Ti0.01O3, and Y0.94Mn1.05Ti0.01O3) by scanning electron microscopy (SEM) in secondary electron emission mode with a small acceleration voltage. Using SEM at an acceleration voltage of 1.0 kV otherwise homogenous surface charging effects are reduced, polarization charges can be observed and polarization directions (±Pz) of the ferroelectric domains in the polycrystalline thin films can be identified. Thin films of different chemical composition have been deposited by pulsed laser deposition on Pt/SiO2/Si structures under otherwise same growth conditions. Using SEM it has been shown that different charged domain density networks are existing in polycrystalline yttrium manganite thin films.

Disturbing-Free Determination of Yeast Concentration in DI Water and in Glucose Using Impedance Biochips.

Deionized water and glucose without yeast and with yeast (Saccharomyces cerevisiae) of optical density OD600 that ranges from 4 to 16 has been put in the ring electrode region of six different types of impedance biochips and impedance has been measured in dependence on the added volume (20, 21, 22, 23, 24, 25 µL). The measured impedance of two out of the six types of biochips is strongly sensitive to the addition of both liquid without yeast and liquid with yeast and modelled impedance reveals a linear relationship between the impedance model parameters and yeast concentration. The presented biochips allow for continuous impedance measurements without interrupting the cultivation of the yeast. A multiparameter fit of the impedance model parameters allows for determining the concentration of yeast (cy) in the range from cy = 3.3 × 107 to cy = 17 × 107 cells/mL. This work shows that independent on the liquid, i.e., DI water or glucose, the impedance model parameters of the two most sensitive types of biochips with liquid without yeast and with liquid with yeast are clearly distinguishable for the two most sensitive types of biochips.

Spectroscopic ellipsometry and magneto-optical Kerr effect spectroscopy study of thermally treated Co60Fe20B20 thin films.

We report the optical and magneto-optical properties of amorphous and crystalline Co60Fe20B20 films with thicknesses in the range of 10 nm to 20 nm characterized using spectroscopy ellipsometry (SE) and magneto-optical Kerr effect (MOKE) spectroscopy. We derived the spectral dependence of the dielectric tensor from experimental data for samples prior and after annealing in vacuum. The features of the dielectric function can be directly related to the transitions between electronic states and the observed changes upon annealing can be ascribed to an increase of the crystalline ordering of CoFeB.

Ferrocenyl-Pyrenes, Ferrocenyl-9,10-Phenanthrenediones, and Ferrocenyl-9,10-Dimethoxyphenanthrenes: Charge-Transfer Studies and SWCNT Functionalization.

The synthesis of 1-Fc- (3), 1-Br-6-Fc- (5 a), 2-Br-7-Fc- (7 a), 1,6-Fc2 - (5 b), 2,7-Fc2 -pyrene (7 b), 3,6-Fc2 -9,10-phenanthrenedione (10), and 3,6-Fc2 -9,10-dimethoxyphenanthrene (12; Fc=Fe(η5 -C5 H4 )(η5 -C5 H5 )) is discussed. Of these compounds, 10 and 12 form 1D or 2D coordination polymers in the solid state. (Spectro)Electrochemical studies confirmed reversible Fc/Fc+ redox events between -130 and 160 mV. 1,6- and 2,7-Substitution in 5 a (E°'=-130 mV) and 7 a (E°'=50 mV) influences the redox potentials, whereas the ones of 5 b and 7 b (E°'=20 mV) are independent. Compounds 5 b, 7 b, 10, and 12 show single Fc oxidation processes with redox splittings between 70 and 100 mV. UV/Vis/NIR spectroelectrochemistry confirmed a weak electron transfer between FeII /FeIII in mixed-valent [5 b]+ and [12]+ . DFT calculations showed that 5 b non-covalently interacts with the single-walled carbon nanotube (SWCNT) sidewalls as proven by, for example, disentangling experiments. In addition, CV studies of the as-obtained dispersions confirmed exohedral attachment of 5 b at the SWCNTs.

P-N Junction-Based Si Biochips with Ring Electrodes for Novel Biosensing Applications.

In this work, we report on the impedance of p-n junction-based Si biochips with gold ring top electrodes and unstructured platinum bottom electrodes which allows for counting target biomaterial in a liquid-filled ring top electrode region. The systematic experiments on p-n junction-based Si biochips fabricated by two different sets of implantation parameters (i.e. biochips PS5 and BS5) are studied, and the comparable significant change of impedance characteristics in the biochips in dependence on the number of bacteria suspension, i.e., Lysinibacillus sphaericus JG-A12, in Deionized water with an optical density at 600 nm from OD600 = 4-16 in the electrode ring region is demonstrated. Furthermore, with the help of the newly developed two-phase electrode structure, the modeled capacitance and resistance parameters of the electrical equivalent circuit describing the p-n junction-based biochips depend linearly on the number of bacteria in the ring top electrode region, which successfully proves the potential performance of p-n junction-based Si biochips in observing the bacterial suspension. The proposed p-n junction-based biochips reveal perspective applications in medicine and biology for diagnosis, monitoring, management, and treatment of diseases.

Synthesis of Mg and Zn diolates and their use in metal oxide deposition.

The synthesis of complexes [M(OCHMeCH2NMeCH2)2] (5, M = Mg; 7, M = Zn) is described. Treatment of MeHNCH2CH2NMeH (1) with 2-methyloxirane (2) gave diol (HOCHMeCH2NMeCH2)2 (3), which upon reaction with equimolar amounts of MR2 (4, M = Mg, R = Bu; 6, M = Zn, R = Et) gave 5 and 7. The thermal behavior and vapor pressure of 5 and 7 were investigated to show whether they are suited as CVD (= chemical vapor deposition) and/or spin-coating precursors for MgO or ZnO layer formation. Thermogravimetric (TG) studies revealed that 5 and 7 decompose between 80-530 °C forming MgO and ZnO as evidenced by PXRD studies. In addition, TG-MS-coupled experiments were carried out with 7 proving that decomposition occurs by M-O, C-O, C-N and C-C bond cleavages, as evidenced from the detection of fragments such as CH4N+, C2H4N+, C2H5N+, CH2O+, C2H2O+ and C2H3O+. The vapor pressure of 7 was measured at 10.4 mbar at 160 °C, while 5 is non-volatile. The layers obtained by CVD are dense and conformal with a somewhat granulated surface morphology as evidenced by SEM studies. In addition, spin-coating experiments using 5 and 7 as precursors were applied. The corresponding MO layer thicknesses are between 7-140 nm (CVD) or 80 nm and 65 nm (5, 7; spin-coating). EDX and XPS measurements confirm the formation of MgO and ZnO films, however, containing 12-24 mol% (CVD) or 5-9 mol% (spin-coating) carbon. GIXRD studies verify the crystalline character of the deposited layers obtained by CVD and the spin-coating processes.

Electrical Conductivity Modeling of Graphene-based Conductor Materials.

Graphene-based conductors such as films and fibers aim to transfer graphene's extraordinary properties to the macroscopic scale. They show great potential for large-scale applications, but there is a lack of theoretical models to describe their electrical characteristics. We present a network simulation method to model the electrical conductivity of graphene-based conductors. The method considers all of the relevant microscopic parameters such as graphene flake conductivity, interlayer conductivity, packing density, and flake size. To provide a mathematical framework, we derive an analytical expression, which reproduces the essential features of the network model. We also find good agreement with experimental data. Our results offer production guidelines and enable the systematic optimization of high-performance graphene-based conductor materials. A generalization of the model to any conductor based on two-dimensional materials is straightforward.

ß-Ketoiminato-based copper(ii) complexes as CVD precursors for copper and copper oxide layer formation.

The synthesis of ketoiminato copper(ii) complexes [Cu(OCRCHC(CH3)NCH2CH2X)(µ-OAc)]2 (X = NMe2: 4a, R = Me; 4b, R = Ph. X = OMe: 5, R = Me) and [Cu(OCRCHCMeNCH2CH2NEt2)(OAc)] (6, R = Me) from RC(O)CHC(CH3)N(H)CH2CH2X (X = NMe2: 1a, R = Me; 1b, R = Ph. X = NEt2: 1c, R = Me. X = OMe: 2, R = Me) and [Cu(OAc)2·H2O] (3) is reported. The molecular solid-state structures of 4-6 were determined by single crystal X-ray diffraction studies, showing that 4a,b and 5 are dimers which are set up by two [{Cu(µ-OAc)L}] (L = ketoiminato ligand) units featuring a square-planar Cu2O2 core with a distorted square-pyramidal geometry at Cu(ii). In contrast, 6 is monomeric with a tridentate-coordinated OCMeCHCMeNCH2CH2NEt2 ligand and a σ-bonded acetate group, thus inducing a square-planar environment around Cu(ii). The thermal behavior of all complexes was studied by TG (Thermogravimetry) and DSC (Differential Scanning Calorimetry) under an atmosphere of Ar and O2. Complex 4b shows the highest first onset temperature at 213 °C (under O2) and 239 °C (Ar). PXRD studies confirmed the formation of CuO under an atmosphere of O2 and Cu/Cu2O under Ar. TG-MS studies, exemplarily carried out with 4a, indicate the elimination of the ketoiminato ligands with detectable fragments such as m/z = 15, 28, 43, 44, 45, and 60 at a temperature above 250 °C. Vapor pressure measurements displayed that 5 shows the highest volatility of 3.6 mbar at 70 °C (for comparison, 4a, 1.4; 4b, 1.3; 6, 0.4 mbar) and hence 4a and 5 were used as MOCVD precursors for Cu/Cu2O deposition on Si/SiO2 at substrate temperatures of 450 °C and 510 °C. The deposition experiments were carried out under an atmosphere of nitrogen as well as oxygen. The as-obtained layers were characterized by SEM, EDX, XPS, and PXRD, showing that with oxygen as the reactive gas a mixture of metallic copper and copper(i) oxide without carbon impurities was formed, while under N2 Cu films with 53-68 mol% C contamination were produced. In a deposition experiment using precursor 5 at 510 °C under N2 a pure copper film was obtained.

Polymer-based doping control for performance enhancement of wet-processed short-channel CNTFETs.

The electrical transport properties of short-channel transistors based on single-walled carbon nanotubes (CNT) are significantly affected by bundling along with solution processing. We report that especially high off currents of CNT transistors are not only related to the incorporation of metallic CNTs but also to the incorporation of CNT bundles. By applying device passivation with poly(4-vinylpyridine), the impact of CNT bundling on the device performance can be strongly reduced due to increased gate efficiency as well as reduced oxygen and water-induced p-type doping, boosting essential field-effect transistor performance parameters by several orders of magnitude. Moreover, this passivation approach allows the hysteresis and threshold voltage of CNT transistors to be tuned.

Magnesium ß-ketoiminates as CVD precursors for MgO formation.

The synthesis and characterization of bis(ketoiminato)magnesium(ii) complexes of composition [Mg(OCR2CH2CHR1NCH2CH2X)2] (X = NMe2: 3a, R1 = R2 = Me; 3b, R1 = Me, R2 = Ph. X = OMe: 3c, R1 = R2 = Me) are reported. Complexes 3a-c are accessible by the reaction of C(O)R2CH2CHR1N(H)CH2CH2X (X = NMe2: 1a, R1 = R2 = Me; 1b, R1 = Me, R2 = Ph. X = OMe: 1c, R1 = R2 = Me) with Mg n Bu2. The structure of 3b in the solid state was determined by a single crystal X-ray diffraction study, confirming that the Mg(ii) ion is hexa-coordinated by two ketoiminato ligands, while each of the latter coordinates with its two N- and one O-donor atom in an octahedral MgN6O2 coordination environment in the OC-6-33 stereo-isomeric form. The thermal behavior of 3a-c was studied by TG and DSC under an atmosphere of Ar and O2 respectively. The respective Me-substituted complexes 3a,c decompose at lower temperatures (3a, 166 °C; 3c, 233 °C) than the phenyl derivative 3b (243 °C). PXRD studies indicate the formation of MgO. Additionally, TG-MS studies were exemplarily carried out for 3a, indicating the release of the ketoiminato ligands. Vapor pressure measurements were conducted at 80 °C, whereby 3a,c possess with 5.6 mbar (3a) and 2.0 mbar (3c) significantly higher volatilities than 3b (0.07 mbar). Complexes 3a-c were used as MOCVD precursors for the deposition of thin MgO films on silicon substrates. It was found that only with 3a,c thin, dense and rather granulated MgO layers of thicknesses between 28-147 nm were produced. The as-deposited MgO layers were characterized by SEM, EDX, and XPS measurements and the thicknesses of the as-deposited layers were measured by Ellipsometry and SEM cross-section images. Apart from magnesium and oxygen a carbon content between 3-4 mol% was determined.

Length separation of single-walled carbon nanotubes and its impact on structural and electrical properties of wafer-level fabricated carbon nanotube-field-effect transistors.

For an industrial realization of devices based on single-walled carbon nanotube (SWCNTs) such as field-effect transistors (FETs) it becomes increasingly important to consider technological aspects such as intrinsic device structure, integration process controllability as well as yield. From the perspective of a wafer-level integration technology, the influence of SWCNT length on the performance of short-channel CNT-FETs is demonstrated by means of a statistical and comparative study. Therefore, a methodological development of a length separation process based on size-exclusion chromatography was conducted in order to extract well-separated SWCNT dispersions with narrowed length distribution. It could be shown that short SWCNTs adversely affect integrability and reproducibility, underlined by a 25% decline of the integration yield with respect to long SWCNTs. Furthermore, it turns out that the significant changes in electrical performance are directly linked to a SWCNT chain formation in the transistor channel. In particular, CNT-FETs with long SWCNTs outperform reference and short SWCNTs with respect to hole mobility and subthreshold controllability by up to 300% and up to 140%, respectively. As a whole, this study provides a statistical and comparative analysis towards chain-less CNT-FETs fabricated with a wafer-level technology.