Scanners and steppers for advanced nodes cost $200 million to $400 million per EUV system, and the small set of global suppliers run multi-quarter lead times, so procurement strategy now drives fab schedules as much as process engineering does [S1].
The procurement question is no longer "which tool" but "which tool, in which lifecycle state, against which node roadmap," and the answer splits cleanly into new-tool, refurbished, and alternative-lithography tracks with very different cost, risk, and qualification profiles [S1][S2][S4].
Match the Wavelength and NA to the Node, Not the Marketing Brochure
Every lithography decision starts with the wavelength stack: i-line at 365 nm covers mature and power-device nodes, KrF at 248 nm spans roughly 150 nm to 110 nm features, ArF dry and ArF immersion at 193 nm cover the 65 nm to 7 nm range, and EUV at 13.5 nm is required below 7 nm [S4]. Numerical aperture (NA) is the second axis: an ASML PAS 5500/150 i-line stepper delivers 180 nm resolution at NA 0.57, while an ASML XT:1700Fi immersion scanner reaches 45 nm resolution at NA 1.20, and a Nikon NSR-S635E pushes that to NA 1.30 [S4].
Choosing by node alone is a common mistake; production volume, overlay budget, and reticle cost per layer swing the answer. A Canon FPA-3000iW at NA 0.48 and 280 nm resolution is the right tool for 120 WPH power-device and MEMS lines, whereas the same fab trying to mix in 180 nm analog would find the ASML PAS 5500/150 with 25 nm overlay the only defensible option [S4]. Equipment spec maps for adjacent process tools, such as the coater/developer line selection in wet vs dry separator line equipment specs, use the same node-to-spec discipline and translate directly to scanner procurement.
New vs Refurbished: The 40 to 70 Percent Capex Swing
Refurbished lithography systems typically trade at 30 to 60 percent of new-tool list, with pricing on the secondary market for a 2019-vintage immersion scanner commonly landing in the $25M-$60M band versus $200M-$400M for a new EUV system [S1][S6]. The supply chain argument is straightforward: lead times of 12 to 18 months for new EUV tools force fabs to either delay ramp or accept used equipment that can be installed in 3 to 6 months [S1].
The trade-off is ownership risk. ASML systems hold resale value best because the installed base is the largest and service network the deepest, so a used ASML PAS 5500/250 KrF scanner at 110 nm resolution and 20 nm overlay is usually the lowest-risk refurbished pick [S4]. Canon FPA-5000ES3 KrF tools offer slightly higher throughput (130 WPH versus 120 WPH) at marginally looser resolution (130 nm) and benefit from Canon's strong Asia-Pacific service network, which matters when field-service response time is a hidden procurement line item [S4]. Nikon NSR-S307E units appear less often on the secondary market but can carry pricing advantages of 10 to 20 percent for buyers willing to qualify a thinner parts ecosystem [S4].
Five Inspection Points That Decide Whether a Used Tool Is a Bargain or a Write-Off
Universal inspection points apply regardless of supplier, and skipping them is the single most expensive mistake in used-tool procurement [S4]. First, the optical column: check lens contamination or damage, illumination uniformity across the field, and (for DUV systems) laser power and stability against the original factory acceptance data [S4]. Second, the wafer stage: measure positioning accuracy and repeatability, inspect encoder and interferometer condition, and confirm smooth motion without stiction at low velocity [S4].
Third, request complete historical maintenance records, including lamp or laser hours, column-replacement dates, and stage overhaul history, because a tool that has had a recent column swap is worth more than one with a pristine log but 40,000 lamp hours [S4]. Fourth, validate the chiller, gas, and environmental enclosure because vibration isolation failures and amine-filter saturation on chemical air handlers routinely show up as overlay drift six to twelve months after install [S4]. Fifth, run a site acceptance test on a witness wafer using a production recipe before signing the bill of sale; this is the only way to catch handshake problems between the tool and the existing track, and the same qualification rigor used in fuse selection spec mapping applies in spirit: verify against the worst-case operating envelope, not the catalog number.
EUV Procurement: Treat the Tool as a 10-Year Ecosystem Contract
Selecting EUV lithography equipment is one of the most consequential capital decisions a fab can make, and the specification is only a fraction of the total evaluation; ecosystem readiness and strategic partnership carry equal weight [S3]. An EUV scanner at $200M-$400M is typically 25 to 35 percent of a single tool set's cost when pellicles, reticle stock, source power upgrade paths, and the EUV-grade photoresist program are included [S1].
The practical procurement gate is therefore: confirm the source power roadmap (250 W to >500 W class), confirm the reticle-handling automation interface, and confirm the scanner's ability to support high-NA upgrade kits in place [S3]. The same multi-facet evaluation logic appears in industrial procurement for dry-type transformer sizing and selection, where the transformer spec is only one input to a much larger site-power contract. For most non-leading-edge fabs, the right answer is not to buy EUV but to qualify ArF immersion with multi-patterning, which can extend a 193 nm dry tool to effective 28 nm-class features at one-quarter the per-tool cost of an EUV system [S4].
Alternative Lithography: Nanoimprint and Mask Aligners for Specialty Volumes
For photonics, LED, MEMS, and certain memory or patterned-media devices, nanoimprint lithography (NIL) and proximity mask aligners displace projection steppers entirely. SUSS MicroTec MA6 and BA6 mask aligners, for example, appear on the secondary market near $350,000 per set and serve 0.5 µm to 5 µm features without the NA and overlay burden of a stepper [S5]. Throughput for these tools runs 20 to 60 WPH depending on exposure mode and substrate handling, which is acceptable for sub-1,000 wafer-start-per-month specialty lines but disqualifies them for high-volume logic [S5].
For buyers who need sub-100 nm features without the EUV price tag, SUSS MicroTec nanoimprint tools reach below 20 nm resolution on a 4-inch to 8-inch wafer at 4 to 10 WPH throughput, with the qualification gate being mask defectivity and template lifetime rather than source power [S5]. The procurement verdict: nanoimprint and proximity aligners are a defensible primary lithography choice for photonics, biotech, and certain power-device lines, but they are not interchangeable with projection steppers, and they require their own template-supply and defect-metrology infrastructure that must be priced into the contract [S5][S2].
Lead-Time Economics and the Strategic Procurement Loop
Lead times for new lithography tools typically run 12 to 18 months from purchase order to first wafer through, and that window is the binding constraint on fab ramp curves, not the equipment budget itself [S1]. The strategic procurement loop therefore has three checkpoints: (1) lock the node roadmap and wavelength mix 24 to 36 months before tool delivery, (2) place conditional purchase orders with a refundable deposit to secure a slot in the supplier build queue, and (3) pre-stage the facility (vibration, amine filtration, power, and chilled water) so the tool can install in 8 to 12 weeks once on site, rather than the 6 to 9 months lost when the fab is not ready [S1].
The secondary effect of long lead times is that used-tool inventories tighten as more fabs try to bridge their EUV delivery gap with refurbished ArF immersion, and prices for late-model Nikon NSR-S635E and ASML XT:1700Fi systems firmed 10 to 20 percent year-over-year through 2025 [S4]. Procurement teams watching 2026 delivery slots should track two signals: ASML's quarterly book-to-bill and lead-time guidance, and the volume of late-vintage immersion tools being de-installed from memory fab conversions in South Korea and Japan, since both lead and lag the secondary-market price by roughly one quarter [S1][S4].
Final procurement gate: confirm wavelength-NA match to node, lock the new-versus-refurbished decision against a 12-month lead-time horizon, complete the five-point used-tool inspection if the tool is secondary, and price the full ecosystem (source, reticle, photoresist, facility) before signing; the next trackable signal is ASML's Q4 2026 delivery and lead-time guidance.
Spec-level background on the components involved: linear guide, crossed roller guide, and ndt equipment.