1 Bewertungen

Semiconductors

Understanding and characterizing nanostructures plays an essential role in the unprecedented technology developments in areas such as information processing, full color displays, and new sensor technologies, to name but a few. In this section, we show how different measurement solutions from Anton Paar are important for the technical progress of our time. The solutions include: characterization of particle size and the study of surface zeta potential to improve the chemical-mechanical polishing process; the use of atomic force microscopy to study metallization in the chip production process or to investigate the surface properties of micro-lenses in CMOS sensors; and analysis of nanopatterned surfaces with grazing-incidence small-angle X-ray scattering (GISAXS).

Click here to find out more about semiconductor metrology in general

Analysis of metallization for the chip fabrication process

Introduction

Metallization is one of the multiple steps that deposits a thin metal layer as a conductive pattern to connect various components in a chip, or produces a so-called bonding pad for the bonding of wire leads from a package to a chip. Besides copper, aluminum is the most widely used material for metallization. Aluminum has good conductivity, good adhesion on SiO2, low electrical resistance and good contact with wire bond. It is easy to deposit and also easy to pattern in single deposition and etching processes. Polyimide, a synthetic polymer, is very commonly used as an adhesive or as an insulating layer in various steps during the chip fabrication process. It is light, flexible and has good mechanical and adhesive strength in relation to materials commonly used in microelectronics. Polyimide also has excellent electrical, chemical and heat resistance. In this report we use the Tosca 400 AFM to characterize both polyimide and metallized aluminum thin films.

Experimental

Both kinds of thin layers were measured by using the tapping mode under ambient conditions. The measurements were performed in an ambient atmosphere and supported by an active anti-vibration isolation table and an acoustic enclosure. Both components minimize external sources of noise and are part of the recommended setup of the Tosca series AFM. One of the key control parameters for the coating quality is the surface roughness which has a large influence on the adhesive strength between layers. The surface roughness was calculated according to the ISO25178 standard using Tosca Analysis software. AFM measurements were also performed on several other randomly selected locations on the sample surface.

Results and discussion

Sq, the root mean square height, is often referred to as the surface roughness. The surface roughness of polyimide thin film is 26.3 nm according to ISO 25178. The surface topography showed no defects on the thin film and the surface roughness was very similar in different locations, indicating a good quality of the deposition of polyimide thin film. Figure 1 shows the surface topography of the polyimide thin film and the corresponding surface height parameters calculated. Figure 2 shows the surface structure of the aluminum thin film after metallization. Grains with different sizes varying from several µm to several tens of µm were observed. A zoomed scan clearly showed the grain boundaries. The surface roughness (Sq) of the aluminum thin film equaled 10 nm.

Height parameters (ISO 25178)

Sq

26.3 nm

Sa

20.6 nm

Sp

92.5 nm

Sv

79.8 nm

Figure 1. Surface topography of polyimide thin film and the height parameters according to ISO 25178

Figure 2. Surface topography of aluminum thin film after metallization process

Additional information

Instruments:

Application report: 

Characterization of nanoparticles to improve the CMP process

Introduction

Chemical mechanical polishing (or planarization) is the most popular technique for removing the surface irregularities of silicon wafers. In the CMP process, the use of slurries with abrasive particles (e.g. silica) allows achievement of highly smooth and planar material surfaces by combining chemical and mechanical actions. The role of the nanoparticles in the CMP process is to remove impurities without adhering to the silicon wafer. Moreover, the particle size has to be monitored because it affects the rate of removal and wafer defects. The particle size, as well as the electrostatic repulsion between slurry particles and the wafer surface, thus drive the success of a CMP process.

Experimental

Sample: A commercial slurry of silica particles was used. The pH dependence of the particle size and zeta potential was automatically recorded by using the automatic titration system of the Litesizer 500. For the surface zeta potential measurements, a 6” silicon wafer with a 1000 Å thick silicon oxide (SiO2) layer was cut into pieces of 20 mm x 10 mm and mounted on sample holders of the adjustable gap cell. Streaming current and an aqueous solution of KCl at various ionic strengths (0.001 mol/l and 0.023 mol/l) were selected for the measurements

Characterization techniques: The particle size and zeta potential of the CMP slurry were measured with the Litesizer 500. The streaming current was used to measure zeta potential via SurPASS 3.

Results and discussion

For the estimation of the electrostatic interaction between the SiO2 wafer surface and the silica particles of the CMP slurry, the zeta potential of particles and wafer was compared at the same salinity (0.023 mol/l KCl). The SiO2 wafer exhibited a zeta potential of z » –50 mV, while the CMP slurry at pH 10.8 showed an average zeta potential of z = –45 mV and particle size of 155.4 nm. The example of a SiO2 wafer and a commercial CMP slurry presented in this report suggests the suitable range of pH 6.7-10.8 for an effective CMP process, because the electrostatic repulsion is maximized and this reduces the probability of particle adhesion to the wafer surface.

Figure 1: Comparison of the zeta potential (a) at various pH of an aqueous solution for slurry particles and a silicon oxide wafer. Particle size distribution at pH 10.8.

AFM investigation of CMOS sensor lenses

Introduction

The commonly known CMOS sensor, or complementary metal-oxide-semiconductor active pixel sensor, is widely used in cell phone cameras, web cameras, most digital single-lens reflex cameras (DSLRs), and mirrorless interchangeable-lens cameras (MILCs). In order to obtain clear, sharp and high-resolution pictures, the micro-lenses covering the surface of the CMOS sensor must be manufactured precisely; their surface plays an important role in image quality.

Experimental

The micro-lens array at the surface of a sensor is studied here using atomic force microscopy to investigate size, shape and roughness. A first couple of large scans at 50 µm x 50 µm and 20 µm x 20 µm provided an overview of the array with multiple lenses at the surface of the CMOS sensor. Four lenses were then chosen for further investigation.

Results and discussion

In this investigation, certain irregularities were found at the surface of the lenses. After performing larger scans with 4 lenses in one scan, a single lens scan (Figure 2) was performed by isolating a single lens to study its shape. Following this investigation, a couple of orthogonal profiles were extracted in Tosca Analysis to look at the shape of the lens as illustrated in Figure 3. The first profile (a) revealed an irregularity on one side of the lens not seen in the other direction in profile (b).

Figure 1. Single lens 3D topography

Figure 2. Orthogonal profiles of a single lense

Additional information

Instruments:

Application report:

GISAXS characterization of nanostructured semiconductors

Introduction

Nanopatterned surfaces have wide-ranging applications in different fields. Examples include templates for the deposition of thin films, coating of magnetic films on patterned surfaces, and many others. The key feature of these applications is that the formed nanostructures significantly control macroscopic properties of a material. Ion bombardment can be used for surface cleaning/smoothing of semiconductors, as well as for implementation of dopant atoms. In particular, bombardment with low-energy ions can lead to formation of self-organized periodic patterns, in the size range from a few nanometers, up to µm. These structures can be perfectly analyzed by grazing-incidence small-angle X-ray scattering (GISAXS), which probes a large sample area of surface and surface-near structures and therefore provides averaged, representative results across the sample.

Experimental

GISAXS measurements of semiconductor wafers (germanium and indium arsenide) were performed with the Anton Paar SAXSpoint 5.0 system, which is equipped with a high-precision GISAXS stage. Prior to the measurement, the samples were nanostructured by applying low-energy noble gas ion-induced surface patterning. In addition, the semiconductors were heated to a temperature above the recrystallization temperature which preserves the crystalline structure.

Results and discussion

The scattering reflections originating from the nanostructured surfaces can be recognized from the 2-dimensional GISAXS patterns as displayed in Figure 1.  These patterns were integrated and converted into 1D scattering curves in order to reveal information on the structures, such as periodicity. The two systems studied have a different surface structure. This was further confirmed by atomic force microscopy (AFM) results: the germanium wafer exhibited one main reflection, which corresponded to a periodic structure with around 75 nm d-spacing; in contrast, the indium arsenide sample showed a series of reflections corresponding to a lamellar spacing with a repeating distance of around 110 nm. In summary, GISAXS provided valuable results describing the nanostructured surface created by the applied ion-beam patterning of the semiconductor wafers.

Figure 1. 2D GISAXS patterns, 1D in-plane integrated scattering profiles and AFM images of the germanium (left) and indium arsenide (right) semiconductor wafers.

Additional information

Instruments:

Application reports: