Semiconductor cleaning chemicals are becoming more technically demanding as fabrication advances introduce new materials, integration schemes and process steps. For fab engineers and materials decision-makers, the challenge is not simply removing residues: each clean must address contamination while maintaining compatibility with the materials and structures present at that point in the process.
The market outlook presented in 2019 anticipated continued growth in cleaning products, supported by new applications across front-end-of-line (FEOL), middle-of-line (MOL) and back-end-of-line (BEOL) processing, as well as post-CMP cleaning. The practical implication is clear: cleaning strategies must evolve alongside device architectures, while fabs and suppliers manage formulation complexity, selectivity and supply-chain requirements.
Why semiconductor cleaning requirements are changing
Growth in connected technologies supports demand for semiconductors, but device scaling and new architectures also create manufacturing challenges. Advanced logic, memory extensions and advanced packaging bring additional materials and interfaces into the process. Examples identified in the industry outlook include cobalt, molybdenum and ruthenium, alongside changing integration approaches and replacement chemistries.
These developments affect cleaning across multiple stages. Logic scaling raises concerns around resistance-capacitance (RC) effects in interconnects, while FinFET development, gate-all-around nanowires and high-mobility materials introduce new integration needs. Memory manufacturing faces its own challenges, including 3D NAND growth, DRAM capacitor dielectrics and novel architectures. Patterning requirements, including low-temperature spacers, lithography aids and EUV insertion, add further process complexity.
At the materials level, the outlook highlights needs involving work-function metals, metal hard masks (MHM), cobalt contacts, gap fill, barriers and liners. Each additional material combination can change what a cleaning step must remove—and what it must leave intact.
PERR cleaning: more materials, more selectivity demands
Post-etch residue removal (PERR) has become more fragmented as advanced logic incorporates a wider range of FEOL, MOL and BEOL materials. A clean may encounter combinations of tungsten, cobalt, copper, tantalum nitride, titanium-based materials, aluminum oxide and other layers or plugs. The exact number of cleaning steps and formulations is difficult to define because some chemistries can be tuned for use across more than one material system.
That flexibility matters to fab operations. Maintaining a separate chemistry for every material and integration step can increase process and infrastructure complexity. The industry outlook describes end users seeking formulations that address multiple problems and reduce the number of unique cleans, while still meeting demanding selectivity requirements.
Selectivity is central: a chemistry needs to remove targeted residues without causing unacceptable effects on exposed device materials. Copper, cobalt, titanium nitride, MHM and tungsten are among the materials identified as important considerations. Process teams therefore need to assess cleaning performance in the context of the full stack and integration sequence, rather than evaluating residue removal in isolation.
Post-CMP cleaning expands beyond copper
Post-CMP (PCMP) cleaning has also evolved. Historically, formulated PCMP cleans were concentrated in copper chemical mechanical planarization. The outlook describes growing use of formulated products for cobalt, tungsten and ceria-based polishing applications, in addition to copper.
This expansion follows the broader increase in CMP activity across logic and memory. More polishing applications mean more opportunities for residues and particles to become a process-control concern. Cleaning must be matched to the slurry and polished materials, while supporting the required surface condition after planarization.
Developing formulations for these applications is increasingly difficult. The outlook notes a growing reliance on large multinational chemical suppliers to develop technology-based solutions. For fabs, this makes supplier technical capability and collaboration important considerations as new CMP materials and processes are introduced.
Market growth and supply-chain considerations
The 2019 outlook estimated that the cleaning segments covered in its analysis would grow from $2.378 billion in 2018 to $2.945 billion in 2023, representing a forecast average annual growth rate of 4.4%. It attributed expected growth to new products. The analysis also projected an 8.8% average annual growth rate for formulated products over its forecast period, reflecting the expansion of newer applications alongside established cleaning categories.
These figures are historical forecasts, not current market measurements. Their value for technical decision-makers is the direction they illustrate: new process requirements were expected to support formulated cleaning products, including products for cobalt PERR and cobalt, tungsten and ceria PCMP.
Growth does not remove operational risk. The outlook identifies metrology, defectivity and particle control as challenges likely to increase supply-chain management requirements. It also reported minimal impact from Chinese suppliers at the time and observed that regional suppliers were not meeting leading-edge fab needs for innovative solutions. Supplier mergers and acquisitions were another factor expected to affect the cleaning market.
What fab and supplier teams should consider
As cleaning needs diversify, engineering and procurement teams can use several practical questions to guide process and supplier evaluations:
- Map material exposure: Identify which metals, liners, masks and films are exposed at each cleaning step.
- Define selectivity requirements: Specify both the target residue and the materials that must be preserved.
- Review opportunities to consolidate: Determine whether a formulation can serve multiple applications without compromising process requirements.
- Include post-CMP controls: Assess cleaning as part of the complete polishing and surface-preparation sequence, including particle and defectivity concerns.
- Plan for supply continuity: Consider supplier capabilities and market consolidation when qualifying materials for advanced processes.
Conclusion
Semiconductor cleaning is becoming more closely tied to material innovation and process integration. PERR chemistries must address increasingly varied metal stacks, while PCMP cleaning requirements are extending beyond copper to cobalt, tungsten and ceria-based applications. For fabs, the priority is to balance residue removal, material selectivity and process control while managing the number of distinct chemistries. Continued investment in formulation development and strong supplier collaboration will be important to meeting the cleaning needs of advanced manufacturing.