Transmission Line Model Structures Research Articles

Ultra-low resistance n+GaN contacts for GaN HEMT applications using MOCVD selective area epitaxy in N2 carrier gas

Low resistance n+GaN contact materials were experimentally studied for GaN HEMT applications by selective area epitaxy regrowth on a patterned SiC substrate. Epitaxy was performed by metal organic chemical vapor deposition using 100% H2 or 100% N2 as the carrier gas. Thin film characterization demonstrated that n+GaN grown in N2 carrier gas has a superior morphology with improved crystalline quality to that grown in H2 carrier gas. The results also indicated that the surface morphology of n+GaN grown in N2 carrier gas is less sensitive to mask pattern density and micro-loading effects with Si doping concentrations up to 1 × 1020/cm3. Secondary ion mass spectrometry analysis shows that C and O impurity levels in n+GaN are one order of magnitude lower with N2 carrier gas than with H2. The electrical measurement of transmission line model structures shows an n+GaN sheet resistance of 15 Ω/sq and an Ohmic metal to n+GaN contact resistance of 0.02 Ω-mm for structures grown in N2 carrier gas. These values represent 7.1× and 2.5× improvements compared to H2 carrier gas.

Impact of Ge pre-amorphization implantation on Co/Co-Ti/n+-Si contacts in advanced Co interconnects

Co has been proposed as one of the most promising candidates to replace W or Cu in interconnects, where Co–Ti could be used as a single barrier/liner to prevent Co out-diffusion into dielectric and Si substrate as well as to enhance the adhesion property. In this paper, the material and electrical properties of Co/Co-Ti/n+-Si contacts with different compositions and thicknesses of Co–Ti, with and without Ge pre-amorphization implantation (Ge PAI) were investigated elaborately. Sheet resistance, phase formation, interfacial morphology and element distribution were comprehensively characterized in terms of four-point probe, X-ray diffraction and high-resolution transmission electron microscopy in combination with corresponding energy dispersive X-ray spectroscopy respectively. Meanwhile, the specific contact resistivity (ρc) was extracted using refined transmission line model structures, and it is found that Ge PAI is indeed beneficial in reducing ρc values in Co/Co-Ti/n+-Si contacts. With the assistance of Ge PAI, the lowest ρc value of 8.56 × 10−9 Ω cm2 is accomplished for Co/Co0.65Ti0.35/n+-Si contacts, which is a ∼47.8% reduction in contrast to that without Ge PAI.

Comparison of ohmic contact formation of titanium and zirconium on boron doped diamond

ABSTRACTWith a high affinity to carbon comparable to titanium and an electrically conductive carbide, zirconium has potential to form ohmic contact on boron doped diamond. In this work, formation of ohmic contacts on boron doped diamond using zirconium is studied in comparison to titanium. Boron doped diamond epitaxial layers have been grown by microwave plasma enhanced chemical vapour deposition with various B/C ratio. Circular Transmission Line Model structures were fabricated using standard micro-fabrication technologies. Specific contact resistance of fabricated contacts was determined for different boron concentrations and for various annealing temperatures. Ohmic contacts using zirconium are formed after annealing at 400 °C. Specific contact resistance steadily decreases with high temperature annealing down to a value of ca. 1 mΩ.cm2 after annealing at 700 °C for highly boron doped diamond. In comparison, titanium contact fabricated on highly doped diamond appears not stable under high temperature annealing.

On the Identification of Buffer Trapping for Bias-Dependent Dynamic RONof AlGaN/GaN Schottky Barrier Diode With AlGaN:C Back Barrier

In this letter, we identified a dominant buffer trapping causing a bias-dependent dynamic $R_{\mathrm{\scriptscriptstyle ON}}$ for AlGaN/GaN Schottky barrier diodes (SBDs) fabricated on a C-doped AlGaN buffer as back-barrier. Current transient measurements at various temperatures were performed simultaneously on an AlGaN/GaN SBD and a transmission line model (TLM) structure based on the AlGaN/GaN heterojunction. During the stress of the TLM structure, the two Ohmic contacts are biased at the same voltage forming an equipotential surface while creating a uniform vertical electrical field to induce buffer trapping. We extracted the same dominant trap level of $E_{C}-0.60$ eV from the current transient spectroscopy on both the AlGaN/GaN SBD and the TLM structure after stressing at −100 V, indicating that the increase in the dynamic $R_{\mathrm{\scriptscriptstyle ON}}$ of the diode is due to the electron trapping in the buffer layers. More electron filling of this buffer trap occurs at higher stressing voltages (from −50 to −200 V), which reflects in an enhanced current-transient amplitude at the same time-constant under higher stressing voltages.

Trapping effect evaluation of gateless AlGaN/GaN heterojunction field-effect transistors using transmission-line-model method

An electric-field-dependent trapping effect was observed in our AlGaN/GaN transmission-line-model (TLM) structure. Therefore, we adopted a “gradual voltage stress” scheme to induce a similar trapping effect, that is, a similar sheet resistance variation (ΔRSH) for all intervals of our AlGaN/GaN TLM structure. By measuring the TLM structure under “gradual voltage stress” (stressed TLM measurement), we investigated the trapping effects in gateless AlGaN/GaN heterojunction field-effect transistors. 10 nm of Al2O3 passivation film substantially decreased the voltage-stress-induced ΔRSH, suggesting surface traps were involved. The feasibility and reproducibility of the stressed TLM measurement method was confirmed with repeated passivation.

Evaluation of Transmission Line Model Structures for Silicide-to-Silicon Specific Contact Resistance Extraction

In order to measure silicide-to-silicon specific contact resistance rhoc, transmission line model (TLM) structures were proposed as attractive candidates for embedding in CMOS processes. We optimized TLM structures for nickel silicide and platinum silicide and evaluated them for various doping levels of n- and p-type Si. The measurement limitations and accuracy of the specific contact resistance extraction from the optimized TLM structures are discussed in this paper.

Low‐frequency noise in AlGaN/GaN HEMT structures with AlN thin film layer

AbstractLow‐frequency noise in AlGaN/GaN high electron mobility transistor (HEMT) heterostructures with additional AlN thin barrier layer is investigated. Transmission line model structures with different lengths of the conducting channel formed by polarization effects at the heterointerface of undoped AlGaN/AlN/GaN layers are studied. The measured noise demonstrates an unusual broadening of the generation‐recombination components of the spectra. To explain the noise behaviour of the structure we consider a model taking into account peculiarities of the band structure of the interface with inserted AlN high molar fraction barrier layer. (© 2006 WILEY‐VCH Verlag GmbH & Co. KGaA, Weinheim)

Influence of surface passivation on low‐frequency noise properties of AlGaN/GaN high electron mobility transistor structures

AbstractWe investigated low‐frequency noise in passivated and nonpassivated AlGaN/GaN high electron mobility transistor (HEMT) heterostructures grown on SiC substrate. The heterostructure layers are grown without intentionally doping the barrier material and two‐dimensional gas appears at the interface only due to polarization effects. Transmission line model structures with different lengths of the conducting channel were studied in a wide range of temperatures in a linear regime. The magnitude of the noise and Hooge parameter demonstrate different dependence on the distance between contacts in passivated and non‐passivated structures. Noise reduction in the structure for around one order of magnitude after passivation was revealed and different origins of the noise before and after passivation established. (© 2005 WILEY‐VCH Verlag GmbH & Co. KGaA, Weinheim)

Thermal and electrical properties of Au/B 4C, Ni/B 4C, and Ta/Si contacts to silicon carbide

The role of silicon and boron carbide interface layers on thermal stability and electrical properties of tantalum, gold, and nickel contacts to 6H- and 4H-n-silicon carbide are presented in this report. Thin Ta/Si, Au/B 4C, and Ni/B 4C layers were deposited on SiC using electron-beam evaporation or sputter-deposition techniques. The structures were annealed either in ultra-high vacuum or in N–H ambient, at temperatures ranging from about 400 to 1150 °C. The samples were characterized using atomic force microscopy for surface topography, auger electron spectrometry for depth profiling, and glancing angle X-ray diffraction for microstructure and phase identification analyses. Transmission line model structures for current–voltage measurements and contact resistance evaluation were prepared using photolithography and lift-off techniques. Our results indicate that Ta in the Ta/SiC system decomposes SiC at about 800 °C, forming tantalum carbide with the accumulation of silicon at the TaC/SiC interface. In the Ta/Si/SiC system decomposition of SiC also occurs about the same temperature resulting in the structure TaC/Si/SiC. The Au/B 4C/SiC system appears to be the most thermally stable with the lowest specific contact resistance of about 1×10 −6 Ω cm 2 for samples heat-treated above 1050 °C for 30 min.

Power and temperature dependence of low frequency noise in AlGaN∕GaN transmission line model structures

The low-frequency noise in AlGaN∕GaN transmission line model structures has been investigated for a wide range of ambient temperatures and dissipated powers. A deviation of low-frequency noise from a 1∕f dependence has been observed upon increasing the applied voltage. The effect correlates with the nonlinearity of current-voltage characteristics (CVC). The concept of nonequilibrium 1∕f noise has been introduced to explain 1∕f noise enhancement. A qualitative model connecting non-equilibrium noise with nonlinearity of the CVC via a common origin—nonuniform overheating of the structures—has been suggested. Furthermore, deviations of equilibrium noise from the 1∕f law at low temperatures have been observed and studied.

Equilibrium and non-equilibrium 1/f noise in AlGaN/GaN TLM structures

The low frequency noise of AlGaN/GaN transmission line model (TLM) structures has been investigated in a wide temperature range. Deviation of low frequency noise from 1/f dependence has been observed. The results indicate that the structure become more non-linear upon increasing of power. The observed features have been explained as a result of non-equilibrium condition in TLM structures, equilibrium and non-equilibrium noise cases have been studied. The non-equilibrium fluctuations in gateless TLM structures were analysed taking into account the influence of the positive surface charge, formed by polarization effects on dynamics of non-uniform potential redistribution along the channel.

Improved electrical characterization of Al–Ti ohmic contacts on p-type ion implanted 6H-SiC

This paper deals with the electrical characterization of low resistance Al–Ti 72/28 wt% ohmic contacts to a p-type ion implanted 6H-SiC layer. Transmission line model (TLM) structures were realized on the top of MESA islands defined in this ion implanted layer. A metal scheme composed of Al-1%Si(350 nm)/Ti(80 nm) was deposited by sputtering, photolithography defined and annealed at 1000 °C in Ar for 2 min. TLM structures were measured as a function of the temperature in the range 25–290 °C. The TLM data were mainly analysed by a two-dimensional finite difference simulation tool that takes into account the current crowding effect at the contact periphery. Extracted contact resistivity values fall in the low range of data from the literature. The sheet resistance values computed from the TLM data agreed with those measured using Van der Pauw devices realized next to the TLM structures.

ANNEALING EFFECTS ON CONTACT PROPERTIES OF ALUMINUM DOPED ZINC OXIDE THIN FILMS

Aluminum Zinc Oxide (AZO) thin films are grown on glass substrates by RF Magnetron Sputtering using a single target of zinc oxide (99 wt%) and aluminum oxide (1 wt%) with argon as the plasma. Photolithographic process is then performed on the films in order to obtain a Transmission Line Model structure (TLM) of the metal contact system, namely aluminum and gold. The specific contact resistivity, ρc, of these two metal-semiconductor systems, which will undergo different rapid thermal annealing (RTA) environment, are determined. X-Ray diffraction patterns for these samples are obtained to investigate phase formations or micro-structural changes so as to justify for the differences in specific contact resistivity obtained for these contact systems. The different RTA environment are simulated by purging either nitrogen or argon gas, with a pressure of 40 psi at a temperature of 400°C for 60 s; and annealing in vacuum (10-6 Torr) also at the same temperature and duration. One-dimensional TLM (1D-TLM) measurements are performed on the TLM structures to obtain the specific contact resistivity, ρc. Results show that aluminum contacts on AZO without RTA give the lowest ρc as compared to those in other environment, while gold contacts on AZO annealed in vacuum yield the lowest ρc. Adhesion of aluminum contacts on AZO is good even when subjected to ultrasonic bath test but not true for the case of gold contact, which adheres poorly on AZO films.

Epitaxial layer sheet resistance outside and under ohmic contacts measurements using electrostatic force microscopy

The electrostatic force microscopy (EFM) is used to measure the absolute surface potential on a transmission line model (TLM) structure.The classic electrical measurement taken on the TLM structure has been used to determine the sheet resistance outside the ohmic contact and the contact resistance. Modeling the potential profile along TLM structures has been made using a distributed array of resistances. The sheet resistance of the epitaxial layer under the ohmic contact, the interfacial resistance and the gold layer resistance are taken into account.The measured potential profile using the EFM technique allows to determine the sheet resistance of the epitaxial layer outside and under the ohmic contact which gives the best fit to the model. These values can be determined with an accuracy better than 5%.

Error analysis leading to design criteria for transmission line model characterization of ohmic contacts

The transmission line model (TLM) is a standard method for planar specific contact resistance measurement. Although widely used, the accuracy of a measurement is typically not stated. In addition to contributions from random errors, there can be substantial contributions from systematic errors in typical TLM measurements. In this paper, we develop an analytical model for the experimental uncertainty from the fundamental TLM expressions in order to understand and calculate the uncertainty associated with the specific contact resistance and sheet resistance derived by the TLM method. The experimental uncertainties in measured resistances, together with the pad width and pad spacing, are the dominant contributions to the total uncertainty. Analytical expressions for relative random and systematic uncertainties in contact resistance and sheet resistance are developed in terms of the error contributions and the parameters of the TLM geometry. Expressions for minimum uncertainty, with associated optimum widths and sheet resistances, serve as a basis for the design of TLM structures with minimum uncertainty. The model quantifies the increase in relative uncertainty associated with decreasing contact resistance. Simulations of uncertainty under various sheet resistance, contact resistance, and pad width are implemented and uncertainties are calculated for realistic data sets.

Enhanced transmission line model structures for accurate resistance evaluation of small-size contacts and for more reliable fabrication

Two new transmission line model (TLM) techniques are presented, The first structure, LT-TLM, provides very accurate determination of the transfer length LT (within 5%) and the contact resistivity /spl rho//sub C/ (within 12%) of lateral contacts, as compared to typical errors of 30% and 50%, respectively, for conventional TLM patterns, accurate determination of these quantities is important for the optimal design of small lateral contacts in high-performance devices, such as heterojunction bipolar transistors (HBTs). The second structure, a modified concentric TLM pattern, provides the same results as conventional TLM patterns; however, they do not require mesa isolation, and they can be fabricated extremely reliably. In contrast, conventional concentric TLM patterns are often shorted together or distorted by lift-off metallization processes. Both theoretical analysis and experimental verification with analysis of the accuracy of the extracted data is presented for each technique, demonstrating the advantages of these approaches.

Low frequency noise analysis of LT-GaAs and LT-Al 0.3 Ga 0.7 AsMISFET active layers

Noise spectroscopy has been used to assess the quality of the active layers of LT-GaAs and LT-Al0.3Ga0.7As MISFET devices using transmission line model structures. Noise spectra were studied as a function of insulator thickness. LT-Al0.3Ga0.7As samples and a 500 A thick LT-GaAs sample exhibited 1/f noise. The noise parameter αlatt was found to be of the order 10–4 for these samples. 2000 A LT-GaAs samples exhibited 1/f3/2 noise with 500 Hz corner frequency.

Specific contact resistivity of InGaAs/InP p -isotype heterojunctions

The specific contact resistivity of lattice matched InGaAs/InP p-isotype heterojunctions has been measured through the use of an interface transmission line model structure. The measured resistance values are comparable to or greater than those of the metal/semiconductor interface and depend heavily on the doping and the abrupt or graded nature of the interface.

Statistical modeling of transmission line model test structures. I. The effect of inhomogeneities on the extracted contact parameters

Extraction of electrical contact parameters by transmission line model (TLM) test structures is based on the assumption that the electrical and geometrical parameters of the contacts within a structure are identical. Statistical modeling is presented to show that the usual extraction procedure can lead to large errors in the extracted contact parameters in the case of nonideal TLM structures, even if there is no error in the measured electrical and geometrical parameters. Errors in the extracted parameters as functions of the nominal contact parameters are presented graphically. These graphs can be used to assess the accuracy of the TLM measurements and to optimize TLM test structure design. In the parameter intervals examined, the TLM method works best (yields the fewest errors in all extracted parameters) when the geometrical length of a contact is equal to about two times the electrical transfer length.< <ETX xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">&gt;</ETX>

Statistical modeling of transmission line model test structures. II. TLM test structure with four or more terminals: a novel method to characterize nonideal planar contacts in presence of inhomogeneities

For pt.I see ibid., vol.37, pp.2350-2360, Nov. 1990. A modified transmission line model (redundant TLM or RTLM) test structure (with four or more terminals) and a novel data extraction method are suggested to improve the accuracy of the TLM method. Statistical modeling by both statistical simulation and the method of error propagation shows that with four terminals and independently known sheet resistances between the contacts, the errors of the extracted parameters can be reduced considerably. With the use of five or more terminals, the accuracy of the parameter extraction can also be determined by a single structure, allowing the separation of inhomogeneities within a test structure from the inhomogeneities over the wafer. Experimental data gathered by Kelvin cross-bridge resistors and by traditional and modified transmission line model structures on Al-Si-Cu/TiW/p-n Si contacts clearly show the advantage of the RTLM method over the traditional TLM method. On a wafer with small inhomogeneities, both methods gave consistent results.< <ETX xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">&gt;</ETX>