Refined physical model for understanding the secondary electron doping contrast mechanism

Scanning capacitance microscopy (SCM) is based on a contact-mode variant of atomic force microscopy, which is used for imaging two-dimensional carrier (electrons and holes) distributions in semiconductor devices. We introduced a method of quantification of the carrier concentration by experimentally deduced calibration curves, which were prepared for semiconductor materials such as silicon and silicon carbide. The analytical procedure was circulated to research organizations in a round-robin test. The effectiveness of the method was confirmed for practical analysis and for what is expected for industrial pre-standardization from the viewpoint of comparability among users. It was also applied to other electric scanning probe microscopy techniques such as scanning spreading resistance microscopy and scanning nonlinear dielectric microscopy. Their depth profiles of carrier concentration were found to be in good agreement with those characterized by SCM. These results suggest that our proposed method will be compatible with future next-generation microscopy.

Microscopy is the art and science of visually observing small objects. Electrons are electrically charged subatomic particles typically responsible for the flow of electrical current. Scanning is a method of observation where the focus of attention is moved in a chosen pattern and desired information is gained by assembling observations in sequence. In a scanning electron microscope (SEM), a beam of electrons is scanned in a regular pattern over a small object, and the image of the object appears on a screen by displaying the image points in synchrony with the movements of the electron beam on the object.A familiar example of image scanning is a television receiver, where a picture is created on a screen by rapidly scanning an electron beam across the phosphor face of the television tube. The beam striking the phosphor produces light, which is seen as a picture. The signal that modulates the electron beam in a television set is created in a television camera, which is scanned in the same way that the television screen is scanned. In a SEM, the signals come from special detectors in the microscope, not from an optical camera. Otherwise, the image display in a SEM is very similar to television. Although it is typical for SEM images to be scanned at a slower rate than are television signals, some SEMs display images at the same speed as television.The most common signal for detection in a SEM is the secondary electron signal. Secondary electrons are low‐speed electrons (having energies less than about 50 eV) that result from the interaction of the primary (i.e. beam) electrons with the sample. The secondary electrons are detected, usually by a scintillator/photomultiplier detector, and the resulting signal modulates the display screen.Secondary electrons are popular because they provide a kind of shading that gives a three‐dimensional appearance to an image, making it easy to visualize the structural relationships of parts of the sample (Figure 1).Other signals provide information of a more analytical nature. The signal produced by backscattered electrons is directly (although not exactly proportionally) related to the atomic number of the material being struck by the primary beam. Diffracted electrons provide information about crystal spacings and orientations in the sample. Induced‐current signals are particularly interesting for studying the response of active electronic components.The resolution of a SEM image depends on the quality of the optical system that forms the electron beam, including the nature of the electron source. Most SEMs can typically resolve objects as small as 10 nm. The resolution of specialized SEMs is better than 1 nm. In contrast, the resolution of light microscopes is limited by the wavelength of light to about 500 nm.The magnification of a SEM is not directly related to the optical system, very unlike a light microscope. The magnifying power of a SEM is related to the ratio between the size of the scan on the object compared to the size of the scan on the image display. In this way a SEM functions like a mechanical pantograph that was used, typically in older days, for copying drawings. If the arms of the pantograph were of equal length, the drawing would be copied the same size. If the arms were adjusted to different lengths, the drawing could be magnified or reduced.The primary advantage of a SEM is that it provides familiar‐looking images of surfaces with higher resolution and greater depth‐of‐field than an optical microscope. Although the images are typically familiar looking, the contrast mechanism in a SEM is not the same as the contrast provided by light, and an observer expecting always to interpret a SEM image the same way as a light image could easily be fooled.As the electron beam consists of electrically charged particles, the sample for SEM observation must either be inherently electrically conducting or must be made electrically conducting to yield good images. Otherwise, electrical charge builds up on the sample, the yield of secondary electrons is changed, and the scanning beam can become erratic. Low‐voltage or low‐vacuum SEMs can be used to reduce the charging problem.

We used scanning capacitance microscopy (SCM) and scanning spreading resistance microscopy (SSRM) to determine the drift hole mobility in nanometer thick Si∕Si0.75Ge0.25∕Si quantum wells (QWs), from the carrier concentration profiles (obtained by SCM) and the local resistivity values (obtained by SSRM). A relevant decrease of the hole mobility at room temperature was observed when reducing the QW width from 10 nm down to 1 nm. This effect has been explained in terms of the increasing role of surface scattering at the Si∕SiGe interface for lower QW widths.

In this paper, we present our recent applications of scanning capacitance microscopy (SCM) on specific devices with sampling window as small as 100nm. The dopant related root causes were successfully identified on those devices fabricated with 90nm CMOS technology. The key step in our approach is the development of a sample preparation technique that allows us to precisely x-section through a transistor without being affected by focused ion beam (FIB) artifacts. FIB was used to mark the area of interest with high precision, but it did not expose the devices of interest. Optical microscope and atomic force microscope (AFM) were used to inspect the mechanically polished surface, thus avoiding beam effects from FIB or SEM. In the first application, a doping anomaly was identified in a PFET poly gate, in a single bit failed SRAM cell. In the second application, an asymmetry of a PWell implant profile in a window of 150nm was identified as the cause of leakage in a capacitor array. Our approach may be applied to other scanning probe microscopy (SPM) techniques in the same category, i.e., scanning spreading resistance microscopy (SSRM) or scanning microwave microscopy (SMM).

Secondary electron emission (SEE), a common physical phenomenon, exhibits both prominent applications in various fields and concomitant adverse impacts. Secondary electron yield (SEY) is a parameter for characterizing SEE, which determines the effect of the application and is very sensitive to the surroundings, particularly the magnetic field. Previous studies mainly focused on the influence of the magnetic field on secondary electron trajectories rather than the SEY itself. In this work, a specialized experimental platform was established to measure SEY, through which the influences caused by primary electrons and secondary electron trajectories were eliminated. The research findings demonstrate that as the magnetic field intensity increases, the SEY decreases rapidly and then gradually stabilizes. Meanwhile, a self-developed Monte Carlo simulation program was utilized to verify the above results, and it was concluded that the influence of magnetic field on SEY stems from the increased number of collisions of secondary electrons within the material. These results possess significant guiding significance and reference value for certain application scenarios where SEY suppression is required as well as for the design of multiplier devices aiming to enhance the output by utilizing magnetic fields.

Monte Carlo study of secondary electron emission from SiO2 induced by focused gallium ion beams

Secondary electron emission is developed as a means to quantify and image protein binding to Au surfaces modified with patterned organic thin-film arrays. Alkane thiols were patterned via microcontact printing on gold, and their effects on the secondary electron (SE) yield of the surface, systematically quantified. We show that a self-assembled monolayer (SAM) of hexadecane thiol significantly increases the SE yield over the native gold surface, a yield that increases as a function of alkane chain length (C8-C16). This effect is linearly correlated with the surface potentials and wetting properties of these SAMs. Surface layers comprised of poly(ethylene glycol) (PEG) grafted polyacrylamide polymers behave differently, affecting the SE yield by attenuation according to the polymer thickness. These results demonstrate the relative contributions of factors related to the adsorbate molecular structures that serve to strongly mediate the SE yield, providing a foundation for exploiting them as a quantitative electron imaging probe. The latter capability is demonstrated using a model microfluidic assay in which a series of proteins was spatially addressed to a SAM-based pixel array. The gray scale contrasts seen with protein adsorption are directly correlated with both protein molecular weight and mass coverage. These methods are used in two model protein assay experiments: (1) the measurement of the concentration dependent adsorption isotherm for a model protein (fibrinogen); and (2) the selective recognition of a biotinylated protein layer by avidin. These results demonstrate a unique approach to imaging protein binding processes on surfaces with both high analytical and spatial sensitivity.

Different techniques based on the atomic force microscope (AFM) have been developed in the last few years for the electrical characterization of semiconductor devices. The quality of these measurements strongly depends on the tip which should not only have a small radius of curvature but also a high electrical conductivity. Therefore, the choice of metal as tip material is obvious. We have developed a process scheme for the fabrication of pyramidal metal tips which are integrated into a silicon cantilever. This paper discusses this process in detail and shows how the transition was made from prototyping to batch friction using standard 150 mm silicon wafer technology. Results are presented concerning the application of such probes for two-dimensional carrier profiling of InP and silicon structures using scanning spreading resistance microscopy (SSRM) and scanning capacitance microscopy (SCM). A novel tip configuration called tip-on-tip has also been developed. This concept looks promising for future applications. We demonstrate how such a tip-on-tip configuration can be realized.

The availability of very sharp, wear-proof, electrically conductive probes is one crucial issue for conductive atomic force microscopy (AFM) techniques such as scanning capacitance microscopy, scanning spreading resistance microscopy, and nanopotentiometry. The purpose of this systematic study is to give an overview of the existing probes and to evaluate their performance for the electrical techniques with emphasis on applications on Si at high contact forces. The suitability of the characterized probes has been demonstrated by applying conductive AFM techniques to test structures and state-of-the-art semiconductor devices. Two classes of probes were examined geometrically and electrically: Si sensors with a conductive coating and integrated pyramidal tips made of metal or diamond. Structural information about the conductive materials was obtained by electron microscopy and other analytical tools. Swift and nondestructive procedures to characterize the geometrical and electrical properties of the probes prior to the actual AFM experiment have been developed. Existing contact models have been used to explain variations in the electrical performance of the conductive probes.

Selective area growth (SAG) of gallium nitride (GaN) on silicon (Si) wafers efficiently relaxes the tensile stress that is generated in the GaN layer, when the structure is cooled down to room temperature after the growth. Hence, SAG enables the growth of thicker layers that are capable of operating at higher voltages than those grown in 2D layers. In this study, two GaN layers are grown by SAG on 200 mm-diameter Si(111) wafers by metal organic vapor phase epitaxy for the fabrication of pseudo-vertical p–n diodes. During the growth, the SiH4 precursor flow for the first sample was double than that for the second one. The uniformity of the doping concentration of the layers is investigated by scanning spreading resistance microscopy and the p- and n-type doped regions are examined by scanning capacitance microscopy. A low net doping concentration of 1.4 × 1016 cm−3 is extracted from capacitance–voltage measurements and a destructive breakdown occurs at 700 V for a 90 μm-diameter pseudo-vertical p–n diode. These results show the high potential of the SAG of GaN on Si wafers for vertical power devices.

Experimental observation of FIB induced lateral damage on silicon samples

The extended InGaAs short wavelength infrared (SWIR) detector covers 1.0-2.5 μm wavelength, which plays an important role in weather forecast, resource observation, low light level systems, and astronomical observation and so on. In order to fabricate the high performance extended InGaAs detector, materials structure and parameters were characterized with Scanning Capacitance Microscopy (SCM), Scanning Spreading Resistance Microscopy (SSRM), the spreading of minority carriers and lattice quality were obtained. Mesa etching process, etching damage restoration technique and low temperature passivation technique were used in the fabrication of the extended InGaAs detector. The improvement of material structure and device process was studied by fabricating and measuring different perimeter-to-area (P/A) photodiodes and singledevice, respectively. The dark current density of the extended InGaAs detector obviously was reduced, about 2 nA/cm2 at 170 K. The 512×256 FPAs were fabricated, the peak detectivity and the quantum efficiency of which are 5×1011 cmHz1/2/W and 80%, respectively. The staring image yielded of the 512×256 FPAs is shown, which demonstrates very good imaging quality.

Target material dependence of secondary electron images induced by focused ion beams

Unwilling deformations of secondary electron (SE) images due to charging of an insulating layer on materials is one of important issues for semiconductor industry applications of scanning ion microscopes (SIM). This paper presents a Monte Carlo model of SE emission from SiO₂ in which the charging induced by ion bombardment at the energy range of tens of keV is taken into account. A self-consistent calculation is carried out for the transport of a projectile ion, recoiled material atoms and SEs, the creation of space charges trapped in the material and the resultant electric field in/out the material. Drift motion of trapped charges is calculated as well, where the recombination with a charge of opposite sign is taken into account. Therefore, the evolution of the charging is simulated with successive arrivals of ions. Since the surface voltage is positive due to ejection of SEs and injection of positive ions, some of ejected SEs are drawn back to the surface and can rebound on it; these SEs are unable to produce a net emission. Dynamic changes in the SE yield and surface voltage are compared among He ions, Ga ions and low-energy (<1 keV) electrons, along with the space charge distributions and the in/out electric fields. The net SE yield is decreased during ion bombardment and finally it vanishes, which is different from the case of electron bombardment where the net SE yield (including BSEs) is kept to one due to a balance between coming and outgoing electrons. Even if there is not net emission of SEs, the surface voltage does not reach any steady-state condition but progressively increases due to successive injection of positive ions. The growth rate of the surface voltage depends on both the SE yield with no charging and the spatial distribution of the ions penetrating into the material.

This paper reviews some recent advances in the application of scanning probe microscopy (SPM) electrical characterization techniques to several critical surface and interface issues in SiC technology. High resolution carrier profiling capabilities in SiC of scanning capacitance microscopy (SCM) and scanning spreading resistance microscopy (SSRM) were employed for several applications. These included two-dimensional (2D) imaging of doped regions in SiC devices (to be used as input for device simulations or as a feedback for device processing) and the quantitative evaluation of the electrically active profiles of P (or N) and Al implanted 4H-SiC after high temperature treatments. Furthermore, the electrical modification of the SiO2/4H-SiC interface in MOS devices subjected to post-oxide-deposition treatments in NO or N2O and POCl3 have been investigated, providing quantitative information on the electrical activation of incorporated N or P in the few-nm-thick SiC interfacial region. The lateral homogeneity of metal/SiC interfaces was probed at nanoscale by conductive atomic force microscopy (CAFM), with a special emphasis given to the case of Schottky contacts on 3C-SiC, where the diode behaviour is strongly affected by the high density of electrically active defects in the substrate. Finally, CAFM has been employed to study the current transport in epitaxial graphene (EG) grown on 4H-SiC (0001), revealing the impact of the substrate morphology (terraces and steps or facets) on the local conductivity.