Semiconductors analytical techniques

O. Ganschow. In Analytical Techniques for Semiconductor Materials and Process Characterization. The Electrochemical Society, Pennington, 1990,... 

Semiconductor chemical sensors are characterized by low cost, small size, extra high sensitivity (often unattainable in other analytical techniques) as well as reliability. Moreover, concentration of particles detected is being transformed directly into electrical signal and electronic design of the device is the simplest one which can be arranged for on the active part of the substrate. 

A detailed description of analytical techniques is given in a number of original articles and books [3]. We will focus our interest on comparison of capacities of the mentioned physical and chemical methods with those of semiconductor detectors (SCD) or semiconductor sensors (SCS). These detectors are growing popular in experimental studies. They are unique from the stand-point of their application in various branches of chemistry, physics, and biology. They are capable of solving numerous engineering, environmental and other problems. 

There has been considerable interest in the determination of ions at trace levels as, for example, in applications need high-purity water as in semiconductor processing and the determination of trace anions in amine treated waters. For this investigation, we will define "trace" as determinations at or below 1 pg/1 (ppb) levels. The Semiconductor Equipment and Materials International (SEMI) recommended the use of IC for tracking trace ionic contaminants from 0.025 to 0.5 pg/1 [18]. In addition, the Electric Power Research Institute (EPRI) has established IC as the analytical technique for determining of trace level concentrations of sodium, chloride and sulfate down to 0.25 pg/1 in power plant water [19]. 

The study of semiconductor surface structure and surface chemistry was actually begun several decades ago together with the advent of surface analytical techniques. Many of the earliest surface science studies examined semiconductors. However, for... 

Because of the very important role of impurities in determining semiconductor properties, it is desirable to know their concentrations, at least of the electrically active ones. Of course, the techniques we have discussed in this chapter never make a positive identification of a particular impurity without confirmation by one of the established analytical techniques, such as spark-source mass spectroscopy (SSMS) or secondary-ion mass spectroscopy (SIMS). Once such confirmation is established, however, then a particular technique can be considered as somewhat of a secondary standard for analysis of the impurity that has been confirmed. It must be remembered here that an analytical method such as SSMS will see the total amount of the impurity in question, no matter what the form in the lattice, whereas an electrical technique will see only that fraction that is electrically active. 

SIMS and SNMS are versatile analytical techniques for the compositional characterization of solid surfaces and interfaces in materials research.92-94 As one of the most important applications, both surface analytical techniques allow depth profile analysis (concentration profile as a function of the depth analyzed) to be performed in materials science and the semiconductor industry with excellent depth resolution in the low nm range. For depth profiling in materials science, dynamic SIMS and SNMS using high primary ion beam doses are applied. Both techniques permit the analysis of light elements such as H, , C and N, which are difficult to measure with other analytical techniques. 

Particle-induced X-ray emission (PIXE) is an analytical technique based upon observing fluorescent X-rays. As such, it really is not a nuclear technique since it involves an atomic process, X-ray emission. But the atomic electron shell vacancies that are filled when the X-ray is emitted are created using particle-accelerator beams and one uses typical semiconductor radiation detectors, Si (Li) detectors, to detect the X-rays. 

However, this article is not intended to provide an exhaustive review of the voluminous literature on the application of surface analytical techniques to semiconductor problems. Numerous reviews have been published which have treated various aspects of these applications (1-jj). This article is intended to give an overview, drawing from more recent publications, of the ways in which surface analysis continues to play a vital role in the development and application of the numerous material technologies involved in semiconductor processes. In addition, the need for further development of surface techniques and a summary of the materials problem that do not lend themselves to the available analytical techniques are described. 

Fourier transform infrared spectrophotometry is used widely in the semiconductor industry for the routine determination of the interstitial oxygen content of production silicon wafers. However, the lack of interlaboratory reproducibility in this method has forced the use of ad-hoc calibration methods. The sources of this lack of reproducibility are just beginning to be understood. As investigation of this problem continues and wider acceptance is gained for improved experimental and analytical techniques, a greater degree of reproducibility should be achieved. Furthermore, new standard test methods and standard reference materials being developed by the ASTM (71 ),... 

Neutron activation analysis is one of a small number of methods capable of multi-elemental analysis of subnanogram quantities of contaminants in semiconductors and other materials. Milligram to gram-sized samples of silicon, quartz, graphite, or organic materials are nearly Ideal for the method. The physics of the processes involved is simple, and qualitative identification of components is an Integral part of the quantitative analysis. Except for the need for access to a nuclear reactor, the equipment required is readily available commercially, and is comparable in cost and complexity to that used in other advanced analytical techniques. 

Exactly twenty-three years ago this week a conference was held in Boston on ultrapurification of semiconductor materials. One third of the papers at that conference were devoted to Impurity analyses G). Spark source mass spectrography was the newest and most promising analytical technique available at the time, and I would like to compare the status of SSMS at that time to Its present status. 

---