As semiconductor materials evolve, achieving consistent, high-quality surface finishes is becoming increasingly complex. Advanced substrates such as SiC and GaN demand more effective polishing solutions, placing greater emphasis on the performance and stability of abrasive slurries.
Nanodiamond slurries offer significant potential for improving material removal rates and surface quality. However, their effectiveness depends on precise control of particle dispersion - where even small levels of agglomeration can impact process reliability and yield.
In this application note, we demonstrate how Nano-Tracking Analysis (NTA), using the NanoSight Pro, enables detailed, particle-by-particle insight into nanodiamond slurry behaviour. The study highlights how high-resolution characterisation can support better control of dispersion, helping to optimise Chemical Mechanical Polishing (CMP) performance and reduce the risk of surface defects.
As semiconductor materials evolve, achieving consistent, high-quality surface finishes is becoming increasingly complex. Advanced substrates such as SiC and GaN demand more effective polishing solutions, placing greater emphasis on the performance and stability of abrasive slurries.
Nanodiamond slurries offer significant potential for improving material removal rates and surface quality. However, their effectiveness depends on precise control of particle dispersion - where even small levels of agglomeration can impact process reliability and yield.
In this application note, we demonstrate how Nano-Tracking Analysis (NTA), using the NanoSight Pro, enables detailed, particle-by-particle insight into nanodiamond slurry behaviour. The study highlights how high-resolution characterisation can support better control of dispersion, helping to optimise Chemical Mechanical Polishing (CMP) performance and reduce the risk of surface defects.
In semiconductor manufacturing, both lapping and polishing are used for wafer planarisation. Whilst both processes involve removing material from the surface, they differ significantly in terms of their objectives, precision requirements and mechanisms. Lapping is a ‘rough machining process to shape the surface’, whilst polishing is a ‘final finishing process to achieve a mirror-like surface’1). (Fig. 1)
Polishing is a process that removes surface irregularities and the processed altered layer created by lapping to form a smooth surface at the atomic level. It involves the precise removal of material using a flexible polishing pad and a slurry containing fine abrasive particles such as silica. In current semiconductor manufacturing, CMP (Chemical Mechanical Polishing), which combines physical polishing with chemical reactions, is the mainstream method; by utilising chemical softening, it achieves high flatness and low-damage processing. Whilst colloidal silica and ceria are commonly used for Si wafers, next-generation power device materials such as SiC and GaN are extremely hard and chemically stable; consequently, conventional abrasives result in a low MRR (Material Removal Rate), leading to reduced productivity 2).
Nano-diamond particles are diamond—which possesses the highest hardness—micronized to sizes ranging from a few tens of nanometres to hundreds of nanometres; when appropriately dispersed, they enable high-speed polishing of hard materials. Furthermore, the development of hybrid CMP, which combines physical polishing with chemical reactions, is advancing the simultaneous achievement of both polishing efficiency and surface quality 3).
A typical process is as follows.
When using nanodiamond particle slurry for CMP, the most critical factor is the control of particle dispersion. As nanodiamond particle has high surface energy and tends to agglomerate, the formation of secondary particles causes scratch defects on the wafer. For this reason, it is necessary to evaluate the dispersion state precisely.
Although DLS (Dynamic Light Scattering) is simple and widely used, the information it provides may be insufficient for high-precision applications such as CMP.
In contrast, Nano-Tracking Analysis (NTA) offers the significant advantage of being able to measure nanoparticles directly on a particle-by-particle basis and to quantify the proportion of agglomerated particles as a particle concentration. Furthermore, the high-resolution Finite Track Length Adjustment (FTLA) method incorporated in the NanoSight Pro enables the analysis of distribution structures that are difficult to capture using conventional methods, allowing for more advanced quality assessment.
This application note presents the results of measurements of a nanodiamond particle slurry using the NanoSight Pro and discusses its usefulness and potential applications.
A 1 mm³ sample of nanodiamond particle provided by a collaborative research partner was taken and transferred to a 5 mL vial. Two drops of 0.01% surfactant were added and gently mixed, then made up to 5 mL. This was vortexed for one hour to create a slurry simulating a CMP aqueous dispersion. The resulting stock solution was diluted 1,000 x and stirred for 10 minutes, after which two 50 mL vials were prepared. One was measured immediately (Before US), whilst the other was subjected to 10 minutes of ultrasonic bath treatment (10 min US). The samples were injected into the NanoSight Pro via a syringe pump, and 10 videos were acquired using a 405 nm laser in FTLA mode; these were then combined and analysed.
The particle size distribution measurement results obtained using the NanoSight Pro (Fig. 2) confirmed that the dispersion state of nanodiamond particle was significantly improved by 10 minutes of ultrasonic treatment. Whilst a shoulder was observed on the coarse particle side in the ‘Before US’ sample, it is clear that in the ‘10 min US’ sample, the peak became narrower and sharper, and the particle size distribution converged more uniformly. This is a characteristic result attributable to the high resolution of NTA.
Quantitatively, the results in Table 2 and characteristics are as follows.
Furthermore, visualization of the individual particle’s light spots—a characteristic feature of NTA—directly confirmed an increased number of particles in Fig. 3. This phenomenon provides primary evidence of the dispersion effect caused by aggregated particles. Moreover, the particle concentration estimated by NS Xplorer software indicates approximately a 1.2-fold increase after ultrasonication.
A reduction was observed across all particle size metrics D10, D50 and D90. The most notable change was in D90, which decreased from 216 nm to 187 nm (approx. 13% reduction), indicating substantial disruption of agglomerated particles. As coarse particles are a major cause of scratches in CMP, a reduction in D90 is a key factor as it directly leads to improved process reliability.
The differences observed in the ultrasonic treatment were confirmed as changes of approximately 15% in particle size. This indicates that NTA clearly captured regions that are difficult to detect using conventional methods. The observation of a sharper particle size distribution obtained here is considered to contribute to improved slurry stability, CMP controllability, and reproducibility of MRR.
In summary, NTA has been shown to be extremely effective for evaluating the dispersion of nanodiamond particle slurries, and to contribute to scratch reduction, improved surface quality, and stabilisation of the polishing process in CMP.
We would like to express our deep gratitude to Global Diamond for providing the samples used in this note.
Reliable nanoparticle characterisation is essential for optimising CMP processes and ensuring consistent manufacturing outcomes. By providing direct measurement of particle size distribution and concentration, NTA delivers the level of detail needed to understand and control complex dispersions.
The results presented here show how improved visibility of nanodiamond slurry behaviour can support enhanced process stability, reduced defectivity, and greater reproducibility. For semiconductor manufacturers and researchers, this translates into more confident decision-making and improved process efficiency.
Malvern Panalytical’s NanoSight Pro provides the advanced analytical capability required to meet these challenges - helping you turn particle insight into process control.