Spreading Resistance Technique Research Articles

Influence of rapid thermal processing on carrier concentration in high resistivity silicon

In this study, we report changes in the carrier concentration of high resistivity Si wafers after rapid thermal annealing (RTA) anneals measured by spreading resistance technique. Spreading resistance technique (SRP) profiles clearly show the generation of donor centers with concentrations and depth distributions comparable to those of vacancy‐related centers reported in the literature. Changes of carrier concentrations as a function of RTA temperature, duration, ramp down rate, and subsequent annealing in the 800–1000 °C range are consistent with the earlier literature data. The influence of annealing ambient is also studied. Annealing in pure Ar atmosphere leads to profiles dominated by in‐diffusion of vacancies generated at the surface, while annealing in oxidized ambient results in well‐known profiles controlled by out‐diffusion of vacancies generated in the wafer bulk. Studies of the wafers with different oxygen content show that the concentration of the generated donors is directly proportional to Oi concentration.

Doping compensation for increased robustness of fast recovery silicon diodes

High-power diodes with the radiation enhanced diffusion (RED) of Pd are shown to have much higher ruggedness during the reverse recovery compared to that of the Pt. Anode doping profiles measured by spreading resistance technique after a 10 MeV He implantation with subsequent annealing between 500 and 800 °C reveal different compensation effects between the Pd and Pt. The in-diffusing Pd converts the n-type background doping concentration of N D = 3 × 10 13 cm −3 in the position of radiation defects to that of a p-type with about one order higher concentration. The created low-doped p-layer significantly increases ruggedness of diodes during reverse recovery. In the diodes with the Pt layer, only a modest compensation is observed, a conversion to a p-type layer is missing and robustness is much lower. The DLTS spectra for the Pt and Pd devices show a similar electronic structure and introduction rates of defects at 700 °C, while they differ significantly at 600 and 650 °C both for the majority and minority carriers. It is preliminary suggested that the strong compensation effect after the RED of Pd is caused by a high introduction rate of an acceptor deep level at the lower half of the silicon bandgap.

Diffusion and Electroluminescence Studies of Low Temperature Diffusion of Boron in 3C-SiC

The characteristics of boron diffusion in 3C-SiC at low temperature have been measured using spreading resistance technique and electroluminescence spectroscopy. The coefficient of boron diffusion in the temperature range of 1150 –1250°С has been found to be about 5.5 x 10-11–5.0 x 10-10 cm2/sec and the activation energy of boron diffusion was determined to be about 0.9 –1.15 eV. Electroluminescence spectra of 3C-SiC p-n junction structures showed peaks at 750 and 630 nm due to growth defects and carbon-silicon divacancies respectively.

Atomic transport in germanium and the mechanism of arsenic diffusion

The paper reviews some basic aspects of self- and foreign-atom diffusion in germanium (Ge) and provides new experimental results on the intrinsic and extrinsic diffusion of arsenic (As) in Ge. Experiments of As diffusion in Ge have been performed at temperatures between 640 and 920 °C utilizing a diluted Ge–As alloy as diffusion source. The As profiles were measured by means of the spreading resistance technique. Depending on the As concentration established at the surface, intrinsic and extrinsic As profiles were obtained. Arsenic diffusion in Ge is fully described on the basis of the vacancy mechanism taking into account singly negatively charged As-vacancy pairs. In contrast to a recent study of Vainonen-Ahlgren et al. (Appl. Phys. Lett. 77 (2000) 690), which claims that As diffusion is mediated by neutral and doubly negatively charged vacancies, we show that As diffusion is not sensitive to the properties of vacancies. From our experiments we determine an activation enthalpy of 2.70 eV and a pre-exponential factor of 30 cm 2 s −1 for intrinsic As diffusion in Ge.

A comparative study for profiling ultrathin boron layers in Si

AbstractThe carrier concentration‐depth profiles of ultrathin boron layers in Si, grown by molecular beam epitaxy, are determined by the electrochemical capacitance‐voltage (ECV) and the spreading resistance (SR) profiling techniques. Secondary ion mass spectrometry (SIMS) is employed as a base for the comparison of the results. It has been shown that, under carefully chosen conditions, both ECV and SR techniques are able to resolve ultrathin layers including a delta layer, however ECV match better with the results of SIMS than that of SR. (© 2003 WILEY‐VCH Verlag GmbH & Co. KGaA, Weinheim)

Analysis of the correction factor in the spreading-resistance technique for monitoring epitaxial silicon

The range of applicability of the mixed-boundary-value method for calculating spreading resistance for a homogeneous slab with an effective contact-radius source and backed by a substrate of arbitrary but finite resistivity is investigated. Solutions are presented in terms of the correction factor and the source current density distributions for a slab of varying thickness and with various high-resistivity substrates. The correction factors for different schemes of calculation are correctly obtained by means of introducing an additional current source of opposite sign, but with the same absolute value, located at the drain point contact. This combination of source and drain currents gives the true value for the potential distribution along the surface of the semiconductor. This, in turn, leads to new terms in the equations obtained for the correction factors, which have been omitted in previously published works. A comparison between several schemes of calculation is presented. Within the framework of “uniform” and “variable” current distributions underneath the contact probe, there are two limits for the correction factor. A model based on a combination of these approaches is discussed, and a comparison between the proposed method and the Schumann–Gardner formulation is made.

Kinetics of boron reactivation in doped silicon from Hall effect and spreading resistance techniques

In this work, a series of 13 boron implants were performed into Czochralski silicon substrates with doses of 2×1014–1.6×1015 cm−2 at energies of 10–80 keV. The boron was deliberately clustered with a 750 °C anneal of 10 or 30 min and the electrical activation of the boron implants was determined following a second anneal at 750 or 850 °C with a Hall effect system with certain samples also being analyzed with a spreading resistance technique. Analysis of the reactivation rates allows for the determination of the net energy to boron reactivation to be approximately 3.0 eV assuming the reactivation process is mediated by release of a boron interstitial with a migrational energy of 0.3 eV. This results in a critical binding energy of approximately 2.7 eV from the process limiting the dissolution of the most stable boron-interstitial cluster.

Study of phosphorus implanted and annealed silicon by electrical measurements and ion channeling technique

We investigated the effect of annealing temperature on the electrical activation of phosphorus implanted into silicon. The measurements performed using spreading resistance, four-point probe and ion channeling techniques have allowed us to establish the existence of two domains of variation of the electrical activation (350–700 °C) and (800–1100 °C). The presence of reverse annealing and the annihilation of defects have been put in a prominent position in the first temperature range. It has been shown that in order to achieve a complete electrical activation, the annealing temperature must belong to the second domain (800–1100 °C).

Towards sub-10 nm carrier profiling with spreading resistance techniques

The manufacturing of state-of-the-art electronic devices involves an increasing demand for the accurate determination of ultra-shallow electrical carrier profiles related to the need to monitor the activation of the dopants with reduced thermal budgets. For sub-micron structures (down to 100nm) a qualified conventional spreading resistance probe system is an attractive tool for the reliable measurement of the resistivity (and carrier) depth variations in silicon due to its high geometrical resolution (nm) and high dynamic range (nine orders of magnitude). The spreading resistance (SR) roadmap for future process development (sub-50nm profiles), however, shows that there is a need for a significant reduction of the involved contact size and tip separation, a higher depth resolution (sub-nm) and an improved quantification. The recently introduced scanning spreading resistance microscopy technique resolves some of the involved issues such as the smaller contact size (20–50nm) and the higher geometrical depth resolution (sub-nm) when applied on a bevelled surface. Further developments are, however, needed in the fields of tip configuration, surface preparation and contact modelling to achieve timely all the needs of the SR roadmap. This is expected to lead to a new instrument, the NanoProfiler™, using two small (20–50nm contact size), closely spaced (250nm), conductive tips mounted on an atomic force microscope-based system. The NanoProfilerTM setup can easily achieve Angstrom depth resolution and therefore makes the profiling of sub-10nm structures feasible.

Stripping Hall effect, sheet and spreading resistance techniques for electrical evaluation of implanted silicon layers

The objective of this chapter is to address the electrical characterization of silicon implanted layer. The determination of the electrical activity profiles, in terms of free carriers concentration and mobility, is described using spreading resistance and stripping Hall effect techniques. Several recent results are presented in order to show the importance of such experimental analysis

Profiling of the p–n Junction in Silicon by the Electrochemical Capacitance–Voltage Technique

The formation of p–n junctions by boron implantation in Si was investigated using electrochemical capacitance–voltage (ECV) techniques. An optimized set of parameters for N(x) profiling has led to the choice of bias voltage and determination of the effective dissolution valence of z = 3.3 of the used 0.2 M NH4F · HF electrolyte. The results obtained by ECV were compared with those achieved by the spreading resistance technique and theoretically calculated by SUPREM.

Diffusion of gold in silicon during rapid thermal annealing: Effectiveness of the surface as a sink for self-interstitials

Rapid thermal annealing was used for short-time diffusion experiments of gold in dislocation-free floating-zone silicon of {100} orientation at 1050 °C and 1119 °C. Concentration-depth profiles measured by the spreading-resistance technique are well described within the framework of the kick-out mechanism involving generation of silicon self-interstitials. More specifically, the gold-incorporation rate appears to be controlled by the outdiffusion of excess self-interstitials towards the surfaces. As a special feature, the measurements reveal a continuous increase of the gold boundary concentration which approaches the pertaining solubility limit only after prolonged annealing. This can be interpreted in terms of a limited effectiveness of gold-alloyed {100} silicon surfaces as sinks for self-interstitials. The validity of this interpretation is supported by computer modeling of the experimental data yielding finite values for the self-interstitial surface-annihilation velocity.

Charged Defects at the Interface between Directly Bonded Silicon Wafers

The interface between directly bonded hydrophobic silicon wafers is studied using the spreading resistance technique. It is found that this interface always contains positively charged defects. Possible origins of the defects are discussed on the basis of the morphology of the interface studied previously.

Defects in GaAs after Si indiffusion and annealing: A TEM and CL study

Silicon diffusion was carried out from a thin (50 nm) sputtered film into undoped semi-insulating and Te- or Zn-doped LEC-GaAs at 900 °C for 5 h under various As pressures. Secondary ion mass spectroscopy and spreading resistance technique were used to characterize the Si indiffusion profiles. Lattice defects in highly Si-doped diffusion region were examined as a function of post-diffusion heat treatments (first at 700 °C for 15 min and second at 1000 °C for 30 min) by using transmission electron microscopy of plan-view and cross-sectional samples. Two types of defect were observed in the diffusion region: (i) perfect prismatic loops on {110} planes and of the interstitial type and (ii) Frank faulted loops on {111} planes, also of the interstitial type. A model for the defect formation and the role of Si in the defect generation are discussed in terms of a negative temperature dependence of thermal equilibrium concentrations of V Ga 3−, which are assumed to mediate the Si diffusion under high n-doping conditions. Cathodoluminescence spectra measured at 4 K and 77 K were obtained from the diffusion layer. Si diffusion affects the band-gap luminescence and generates two deep-level emission bands in the 0.9–1.3 eV spectral region. It is suggested that these deep levels are associated with some diffusion-induced defects and defect complexes.

Activation of Implanted Phosphorus Ions into Silicon by Followed Hydrogen Ion Implantations

AbstractThe activation of phosphorus implanted into n-type silicon (100) substrate by followed hydrogen ion(H +) implantation was studied by means of spreading resistance technique(SR), secondary ion mass spectroscopy(SIMS) and transmission electron microscopy(TEM).“The activation ratio” defined by carrier concentration divided by phosphorus concentration was used as a measure of activation of phosphorus. The H + energy, dose and dose rate dependence of activation ratio of phosphorus was investigated.In the case of thermal annealing at 400 °C for 200 minutes the phosphorus atoms were not activated, on the other hand in the case of H + implantation at 400 °C the phosphorus atoms were activated and the activation ratio was increased almost proportionally with the dose. The SIMS data suggested that the depth profile of phosphorus atoms was not changed after activation by H + implantation. The activation ratio was increased with decreasing the dose rate. The TEM data suggested that the density of residual defects was reduced in the case of lower dose rate. The depth profile of activation ratio was similar to that of hydrogen atoms implanted at 20 °C. From these results the activation and recrystalization mechanism is discussed in the view of contribution of elastic collision process between H + ions and substrate atoms.

Three-dimensional concentration profiles of hybrid diffusers in crystalline silicon

The diffusion of ion-implanted Au, Pt, and Zn in crystalline Si has been investigated. The implantation was performed in photolithographically defined areas of the wafer and a spreading resistance technique was used to measure the three-dimensional concentration profiles of the metal atoms after high temperature diffusion anneals. We found that the lateral spread under the mask is larger than the vertical diffusion, especially on the sample side opposite to the implanted diffusion source. All the significant features of the measured profiles can be explained as a consequence of the kick-out mechanism of diffusion for these transition metals. In fact, the peculiar shape of the concentration profiles is determined by the interplay between the influx of interstitial metal atoms and the outflux of silicon self-interstitials generated by the kick-out reaction. Despite the high lateral diffusion, it will be shown that by a suitable combination of implantation fluence and annealing temperature it is possible to limit this lateral spread inside ∼200 μm, while maintaining a high metal concentration in the region under the implanted area. This demonstrates the possibility of using transition metal diffusion to control minority carrier lifetime in a selected area of a semiconductor device.

Computer controlled spreading resistance system for measurement of carrier concentration profiles of silicon structures

A new computer controlled equipment for carrier concentration profile determination of silicon structures by a spreading resistance technique is presented. The required data and plots can easily be obtained by means of a new software package, which has been specially developed for this purpose. Measurements have been performed on different structures in order to verify the suitability of both the measuring system and the software package. The reproducibility has been found to be within 5% for the investigated structures.

Spreading resistance analysis based on the method of regularisation

The method of regularisation is applied to recover, from spreading resistance (SR) measurement data, the dopant profiles of a variety of pn junction structures in which carrier spilling effects are significant. These structures range from those of several microns thick to submicron structures. A comparison is made of two methods of deriving the dopant profile: a direct method which inverts the measured SR profile directly to yield the dopant profile, and an indirect method, which derives the dopant profile from the “on-bevel” carrier profile. It is shown that while the two methods give similar results for the thick structures, the direct method is more accurate for the thin structures, particularly in determining the location of the metallurgical junction. The computation times required by both methods of a 486DX-33 computer are of the order of an hour. Sufficient details of the methods are provided for them to be implementable by users of the spreading resistance technique.

Lateral spread of high energy P and B ions implanted in silicon along the [100] axis and in random direction

B and P ions were implanted along the [100] axis or at 70 tilt angle of silicon wafers covered by a sequence of 10?m wide and 3?m thick SiO 2 stripes. The lateral penetration under the oxide layer, measured by two-dimensional spreading resistance technique, is higher for random than for channeled implants. With increasing the dose the lateral straggling of channeled implants increases due to the disorder introduced in the sample. The results are correlated wvith calculations obtained by the MARLOWE code.

Poisson-based analysis of spreading resistance profiles

This paper discusses carrier diffusion-induced problems that occur in profiling certain silicon structures with the spreading resistance technique. Such structures include p- and n-wells, lightly doped epitaxial layers and very thin epitaxial or diffused layers. Various methods for resolving carrier diffusion problems are discussed. A new method, which we call SRP2, is introduced. In the SRP2 process, the Poisson equation is used to calculate a spreading resistance profile from an assumed dopant profile (usually deduced from conventional spreading resistance analysis or generated by a process simulator such as suprem or predict); the calculated spreading resistance profile is then compared to a measured profile to ‘‘proof-test’’ the assumed dopant profile. Examples of the SRP2 process will be shown for CMOS p- and n-wells and for several other layers, including thin epitaxial and very shallow source-drain layers. The limits of Poisson-based analysis for ultra-shallow layers will be discussed.