November_EDFA_Digital

edfas.org ELECTRONIC DEVICE FAILURE ANALYSIS | VOLUME 22 NO. 4 28 ADVANCES IN CATHODOLUMINESCENCE: RECENT STEPS TOWARD SEMICONDUCTOR FABS AND FA LABS Christian Monachon and Matthew J. Davies Attolight AG monachon@attolight.com EDFAAO (2020) 4:28-33 1537-0755/$19.00 ©ASM International ® INTRODUCTION Cathodoluminescence (CL) is not necessarily a phe- nomenon that everybody can explainoffhand, but it is one that anybody who grew up in the second half of the 20th century has seen. Indeed, all old cathode ray tube (CRT) screens from that era relied on CL—the emission of low- energy light after a material is excited by an electron—to form an image. Cathodoluminescence spectroscopy consists of the analysis of the low energy (0.1-10 eV) photons emitted as an electron excites a material. Generally speaking, what sets cathodoluminescence spectroscopy apart fromother techniques can be summarized as follows: • the incident beam energy is such that luminescence can be excited in a wide variety of materials, from those with small bandgap (< 0.1 eV) up to insulators with bandgap > 10eV • the primary electron beam energy can be tuned to control its penetration depth into the material, enabling depth-sensitive measurements and sub- surface defect localization • the focusing capability of electrons allows for resolu- tions in the nanometer range, much better than any other optical technique • it is inherently correlative with regular electron microscopy imaging techniques A HISTORICAL PERSPECTIVE Cathodoluminescence andCL spectroscopy have been known for quite some time. [1] The first observation of the phenomenon took place in 1874 and first technical uses occurred in the early days of scanning electronmicroscopy (SEM). It first gained popularity for petrographic studies in the 1970s and became more widespread with the growth of the optoelectronics market. Such a slowmaturing of the technology can be attrib- uted to three main factors. First, the technique could only be used in a limited number of materials: the yield of cathodoluminescence can be veryweak, depending on thematerial/class ofmaterial studied, evenwhen the yield is high, it can be very challenging to interpret. The advent of semiconductor materials, especially direct band gap materials such as GaAs- and GaN- based alloys facilitated both the technological use of CL spectroscopy and its interpretation. The second factor is reproducibility; the optical systems as well as the detectors made CL spec- troscopy anunreliable technique especially in comparison with competing technologies, such as photoluminescence spectroscopy. And lastly, it took significant miniaturiza- tion and technological interest in very small features and defects for CL spectroscopy to surpass regular optical techniques and become worthy of more investment. This article reviews SEM-based CL spectroscopy, starting with the technology involved and proceeding to demonstrate its usefulness beyond fundamental research through a few examples of its applications. APPARATUS SCHEMATIC AND MODES OF OPERATION Most of today’s high-resolution CL implementations are SEMs or electron probe micro analyzers (EPMA). In commercially available systems, light collection is per- formed either by inserting amirror between the objective lens of the electron gun (an “add-on” approach), or by using a reflective objective embedded inside the electron column (Fig. 1). For add-on systems, simple parabolic or elliptical