Diamond Turning for Ophthalmic Optics

Diamond Turning for Ophthalmic Optics

Mechanics, Surface Finish, Limitations, and the Special Case of Contact Lenses and Intraocular Lenses

A precision manufacturing review – Heron Operational Excellence

Introduction

Single-point diamond turning (SPDT) is the manufacturing process behind most of the highest-precision optical surfaces produced today, from infrared imaging optics to the molds that shape everyday plastic lenses. Its distinguishing claim is unusual for a subtractive machining process: under the right conditions, it can generate an optical-quality surface directly off the machine, without a separate polishing step. That capability is exactly why it matters so much in ophthalmic device manufacturing, where the surface being cut sits, quite literally, in the optical path of a patient's vision.

This review covers the process itself, the variables that set the achievable surface finish, where the process runs into hard physical limits, and — because it is the reason this review exists — what changes when the workpiece is a soft hydrogel contact lens button or a rigid intraocular lens (IOL) blank rather than a block of aluminum or a silicon mold insert. Citations refer to the numbered reference list at the end; every figure below is drawn from an indexed technical or peer-reviewed source rather than estimated.

Part I — How Single-Point Diamond Turning Works

SPDT is a lathe-based subtractive process in which a workpiece is rotated on a spindle while a single-crystal diamond tool, mounted on a slide with sub-micron positioning resolution, removes material along a precisely controlled path to generate the desired surface geometry — spherical, aspheric, or freeform.^[4,11] What separates it from conventional turning is scale and mechanism. At the depths of cut used in finishing passes — often a few micrometers or less — many materials that are normally brittle can be made to remove material by plastic flow rather than by fracture. This is ductile-mode machining, and it is the reason SPDT can leave a smooth, crack-free surface on materials such as silicon and germanium that would otherwise chip and fracture under a cutting edge.^[1,5]

Achieving and holding ductile-mode conditions is not automatic. It depends on keeping the undeformed chip thickness below a material-specific critical value; one study on silicon reported that ductile-mode behavior required feed rates at or below roughly 0.9 μm per revolution under the tested conditions, with the transition to brittle fracture occurring above that threshold.^[1] Tool geometry plays a direct role here too — negative rake-angle tools generate the high hydrostatic pressure in the cutting zone that suppresses crack initiation and helps maintain the ductile regime in hard, brittle materials.^[1]

The machine, not just the tool

Because SPDT is asked to hold form and finish tolerances at the nanometer level, the machine architecture carries as much of the burden as the cutting tool does. Ultra-precision lathes are built around air-bearing spindles, which run the workpiece on a thin, frictionless film of pressurized air rather than mechanical bearings; porous-graphite air-bearing spindles have been reported with motion error below roughly 12.5 nm across their full speed range, and general air-bearing spindle designs are commonly specified below 50 nm total indicator runout.^[15] The machine base is typically granite or epoxy-granite for vibration damping and dimensional stability, isolated further with pneumatic vibration isolators, and the whole machine usually runs inside a thermally controlled enclosure — environmental variation is often held to roughly ±0.1 °C, since thermal expansion at these tolerances is not a rounding error but a dominant error source.^[15]

Vibration isolation in particular is a site-design problem, not just a machine-design one. Vibration transmitted up through the building floor — from foot traffic, HVAC equipment, nearby machinery, or road and rail traffic outside the building — is treated as a distinct engineering input separate from the vibration the cutting process itself generates. The facilities-design standard used across ultra-precision and semiconductor manufacturing is the Vibration Criterion (VC) curve framework, a set of one-third-octave-band velocity spectra labeled VC-A through VC-E; a VC-E environment, the strictest commonly specified tier, is described as adequate for the most demanding sensitive systems, including long-path, laser-based, small-target equipment — the category ultra-precision diamond turning machines fall into.^[31] Facility engineering studies of machine-tool foundations have specifically modeled how the choice between a floor-mounted foundation and a dedicated isolated concrete block foundation changes vibration transmissibility from the ground into the machine, underscoring that the building slab a diamond turning machine sits on is itself an engineering variable, not a fixed given.^[31]

Part II — What Determines Surface Finish

The starting point for predicting SPDT surface finish is a purely geometric model. For a rounded tool tip cutting at a given feed, the theoretical arithmetic average roughness is approximated by:

Ra ≈ f² / (32 × r)

where f is the feed per revolution and r is the tool nose radius.^[12,26] The relationship is quadratic in feed — doubling the feed rate roughly quadruples the theoretical roughness — while increasing the tool nose radius improves finish. Studies confirm this ordering of importance experimentally: in an optimization of turning parameters, feed rate emerged as the most significant factor governing surface roughness, followed by tool nose radius, followed by spindle speed.^[6]

This geometric model, however, describes a best case, not a guarantee. It assumes a perfectly sharp tool and zero vibration; real diamond-turned surfaces are also shaped by:

• Tool condition — rake angle, cutting-edge waviness (edge sharpness is itself limited by the diamond's own crystal structure at the nanometer scale), and progressive tool wear all directly change the roughness actually generated, independent of the feed/radius geometry. ^[9,11]

• Machine dynamics and floor-borne vibration — spindle error motion, structural vibration, and slide-following error superimpose their own periodic or random error onto the tool path. Vibration is described in the ultra-precision machining literature as the key factor restricting achievable accuracy and surface quality, and a meaningful share of it arrives from outside the machine entirely — transmitted up through the building floor from foot traffic, HVAC plant, and nearby equipment — which is why facility vibration criteria (VC curves) and foundation design are treated as part of the process specification, not an afterthought. ^[10,11,31]

• Thermal drift — because tolerances are held at the nanometer scale, small temperature changes in the tool, spindle, or workpiece produce dimensional drift that shows up directly as form error and finish degradation over a machining run. ^[15]

• Ambient humidity — less discussed than temperature or vibration, but relevant wherever the workpiece itself can absorb moisture. Standard cleanroom practice holds relative humidity in a roughly 45–55% band for general particulate and static control, but hygroscopic workpiece materials can require tighter, application-specific control well outside that general band — a point developed further in Part IV for contact lens lathing specifically. ^[20]

• Material properties — anisotropy, crystallographic orientation, impurities, inclusions, and grain microstructure all influence how a given material actually responds to the cutting edge, meaning two nominally identical materials can diamond-turn to different finishes. ^[6,11]

• Elastic recovery ("spring back") — the tool's finite edge radius elastically deforms material ahead of the cut in addition to removing it; after the tool passes, that material partially recovers toward its original position, which can raise a burr-like ridge and roughen the surface. The magnitude of this recovery scales with the workpiece's elastic modulus relative to its hardness — a fact with direct consequences for soft polymers, discussed in Part IV. ^[10,17]

The reason surface roughness receives this much attention in optics — rather than being treated as a cosmetic finishing detail — is that it directly governs light scatter. Commercially polished glass typically achieves an RMS roughness (Rq) around 1.2 nm, and superpolished optical surfaces can reach 0.1–0.2 nm; a surface with Rq around 25 nm, by contrast, scatters on the order of 10% of incident light, and even a few nanometers of roughness is enough to measurably degrade optical performance at visible wavelengths.^[22]

Part III — Where Diamond Turning Runs Into Hard Limits

Ferrous metals: a chemistry problem, not a hardness problem

It is a common misconception that diamond turning is limited by workpiece hardness. In practice, the sharpest limitation is chemical, not mechanical: diamond tools used to cut ferrous metals (steel, iron, and their alloys) undergo rapid chemical wear through graphitization — the diamond's carbon converts to graphite under the heat, pressure, and catalytic action of the iron at the cutting interface.^[13,14] The effect is dramatic: research on diamond/iron interfaces found the graphitization temperature drops from roughly 5,215 K for natural, uncatalyzed graphitization to around 1,300 K in the presence of iron, meaning the tool edge degrades far faster than its hardness alone would suggest.^[14] This is why direct diamond turning of steel optical tooling is avoided in standard practice, and why mold inserts for many optical and ophthalmic components are instead made from non-ferrous substrates such as aluminum, electroless nickel, or brass.^[13,14]

The ductile-to-brittle line is easy to cross

Hard, brittle materials such as silicon, germanium, and tungsten carbide can be diamond-turned to an optical finish only by staying inside the ductile-mode regime described in Part I — and that regime is narrow. Tool wear itself can trigger a sudden transition from ductile-mode cutting back to brittle fracture mid-cut, degrading a surface that was previously optical-quality without any change in the programmed feed or depth of cut.^[3]

Polymers wear diamond tools too — counterintuitively

Diamond is the hardest known material, so tool wear when cutting a comparatively soft polymer seems like it should be a non-issue. It isn't. Tool wear has been specifically identified as a real problem in diamond turning of polymers, and determining which machining parameters and material properties drive both surface finish and tool wear in polymer turning remains an active area of study rather than a solved problem.^[25]

Practical throughput and cost limits

SPDT is a serial, point-by-point material removal process; it is inherently slower and more expensive per part than replication processes such as injection molding or cast molding. In ophthalmic manufacturing this shapes the whole production strategy: SPDT is typically reserved for making the mold insert or for producing lower-volume, highly customized parts directly, while high-volume standardized parts are replicated from a diamond-turned master rather than cut individually — a distinction that becomes central in Part IV.

Part IV — Contact Lenses and IOLs: What Changes in Medical Device Lathing

Why lathing survives in an industry dominated by molding

Modern soft contact lens manufacturing is dominated by two competing processes: spin-casting, developed by Otto Wichterle at the Czechoslovak Academy of Sciences and commercialized starting in 1966, and lathing, in which an anhydrous cylindrical blank of polymer (a xerogel button) is cut on a lathe and only afterward hydrated into its final soft, hydrated form.^[24,32] Spin-casting is cheaper and better suited to high-volume production of standard spherical lenses, while lathing is the more expensive of the two techniques — but it affords design freedom that spin-casting and simple molding cannot easily match: toric, multifocal, and other non-rotationally-simple or highly customized geometries.^[32] That is precisely why lathing has never disappeared: it remains, by a wide margin, the standard method for rigid gas-permeable (RGP) lenses, and it persists for custom and specialty soft lens designs where a mold-per-parameter approach is not economical.^[19,32]

Quality and regulatory context specific to medical devices

A diamond-turned aluminum telescope mirror and a diamond-turned RGP contact lens button are dimensionally similar problems governed by very different rulebooks. Ophthalmic devices manufactured or finished by lathing fall under ISO 13485 quality management requirements, and materials in direct, extended contact with ocular tissue require biocompatibility testing under the ISO 10993 series along with full material traceability.^[20,30] In practice this means environmental parameters that a general-purpose optics shop might treat as a finish-quality issue — particulate control, temperature and humidity in the lathing environment — are treated in ophthalmic lathing as a patient-safety issue with documented, lot-level records, because incorrect control can affect the finished lens in ways that range from lost batches to patient-facing risks such as microbial contamination.^[20]

Soft (hydrogel) lens materials: machining a shape you haven't made yet

The defining oddity of soft contact lens lathing is that the lens is never machined in the state the patient will wear it. It is cut as a dry, rigid xerogel button, then hydrated afterward, at which point the hydrophilic polymer swells substantially into its final soft, hydrated geometry.^[23,24] That means the dry-state cut geometry has to be deliberately over- or under-sized to land on the correct hydrated dimensions — and the degree of hydration expansion is not perfectly constant or predictable from lens to lens, which makes dry-state dimensional planning a genuine source of manufacturing variability distinct from anything in Parts II or III.^[24] The upside of machining in the dry, rigid xerogel state is precision: modern contact lens lathes are reported capable of cutting xerogel surfaces to tolerances of 8–15 nm, smooth enough that a separate polishing step can be skipped entirely — which matters most for non-spherical (toric, multifocal, custom) geometries that are difficult to polish uniformly in the first place.^[24]

This is also where ambient humidity stops being a general good-practice item and becomes a specified process parameter. Because xerogel and silicone hydrogel lens materials are themselves hygroscopic, atmospheric moisture is a direct competitor to the deliberate, controlled hydration step that happens later — uncontrolled humidity during lathing risks the material absorbing moisture prematurely and unpredictably, undermining the dry-state dimensional planning described above. Patent documentation for silicone hydrogel lens lathing specifies the process should be carried out at relative humidity below about 50%, and preferably below 40%, 30%, or even 25% for tighter control.^[33] Temperature and humidity are controlled together, not separately, for the same underlying reason: published dimensional-stability testing for finished contact lenses is conducted at a specified 21±2 °C and 50±3% relative humidity, precisely so that a lens's dimensions can be verified against a fixed, repeatable environmental reference rather than one that drifts with the season or the building's HVAC cycle.^[34]

Rigid gas-permeable materials: a machinability story

RGP lenses are still, essentially without exception, lathe-cut — it is described in industry technical material as the only way to manufacture them.^[19] Contemporary RGP materials are fluorosilicone acrylates, second-generation polymers that combine a silicone acrylate backbone with fluorinated monomers for better wettability and lower protein deposition than earlier materials.^[27] Their machining behavior is not incidental to that chemistry: silicone acrylate material was found to be a better thermal conductor than PMMA and to warp less than cellulose acetate butyrate (CAB), another earlier rigid lens material, and adding fluoroalkyl methacrylate content has been used specifically to increase pre-hydration hardness and improve machinability.^[27] In other words, RGP material formulation has co-evolved with the lathing process itself — thermal conductivity, warpage resistance, and cuttable hardness are treated as material design targets, not just optical or biological ones. A typical production setup uses a diamond-tipped tool to cut the posterior optical surface into the button, a second diamond tool to reduce the part to final diameter, and modern automated two-axis lathes (the Optoform family being the commonly cited example) built on air-bearing spindles and slides specifically to leave a smooth, polish-free surface directly off the machine.^[19,21]

IOL materials: PMMA, acrylics, and where diamond turning actually sits in the process

It is worth being precise about where SPDT fits in IOL manufacturing, because the answer differs by material. Rigid PMMA IOLs — non-foldable, historically the first widely used IOL material — are the material most directly associated with lathe-based finishing. Foldable IOLs made from hydrophobic or hydrophilic acrylic, by contrast, are predominantly produced by injection molding or cast molding, with diamond turning used upstream to cut the mold inserts that then replicate thousands of lenses; one review describes the traditional foldable acrylate IOL process as combining injection molding with ultraprecision diamond turning in exactly this mold-then-replicate sequence.^[16,17] Where hydrophobic acrylic is diamond-turned directly — for custom IOLs or for research into direct-cut alternatives to molding — process control matters even more than for stiffer materials: one study found that pre-cooling the workpiece significantly reduced both cutting forces and resulting surface roughness, particularly at higher cutting speeds and moderate feed rates, and that surface roughness from conventional room-temperature processing of these materials often remains at or above the micron level rather than reaching the nanometer-level finish needed for optical clarity.^[16,17] For the diamond-turned mold inserts that shape high-volume foldable IOLs, the reported finish target is demanding: surface finishes on optical mold inserts commonly need to reach Ra below roughly 10 nm to keep light scatter and haze in the finished, replicated lens within acceptable limits.^[16]

It is also worth separating two IOL quality issues that are easy to conflate. Surface roughness from machining or molding is one determinant of optical clarity. Glistenings — fluid-filled microvacuoles that form inside some hydrophobic acrylic IOLs once implanted in the eye's aqueous environment — are a different phenomenon, arising from phase separation of water absorbed into the bulk polymer rather than from the cut surface itself; the literature attributes glistening formation to a combination of material composition, manufacturing technique, and packaging, and manufacturing quality improvements have measurably reduced it over time — one comparison found laboratory-induced microvacuole density fell from a mean of about 316 per mm² in IOLs manufactured in 2003 to about 40 per mm² in IOLs manufactured in 2012, an 87% reduction attributed to manufacturing improvements over that period.^[28,29] Newer hydrophobic acrylic formulations with tightly controlled water content (roughly 1.5–4%) have been developed specifically to remain glistening-free while keeping hydrophobic handling characteristics.^[29]

What specifically drives contact lens and IOL surface quality

Pulling the material- and process-specific findings above together, four factors recur as the dominant, evidence-backed drivers of finished surface quality in ophthalmic lathing specifically, beyond the general SPDT factors covered in Part II:

• Elastic modulus and spring-back — soft xerogel and hydrogel polymers have low elastic modulus relative to metals and glasses, which increases elastic recovery behind the cutting edge and makes achieving a given theoretical Ra harder than the feed/tool-radius geometry alone would predict. ^[17]

• State-at-machining versus state-in-use — for soft lenses, the entire process is built around cutting a dry xerogel and predicting how it will expand on hydration, turning dimensional accuracy into a materials-science prediction problem as much as a machining-precision one. ^[24]

• Thermal behavior of the specific polymer — RGP material chemistry has been deliberately tuned for thermal conductivity and warp resistance because heat buildup during cutting directly threatens dimensional accuracy in these materials; the same logic (with cooling applied externally via pre-cooling) shows up in direct machining of hydrophobic acrylic IOL material. ^[16,27]

• Design-driven necessity of lathing over molding — non-rotationally-simple geometries (toric, multifocal, custom irregular-cornea RGP designs) are the reason lathing remains economically justified at all for lower-volume parts, despite being slower and costlier than molding — meaning the process is often chosen because of what it can shape, not because it is the cheaper option. ^[32]

• Environmental control specified as tightly as the machine itself — for hygroscopic lens materials, relative humidity is documented as a controlled lathing-process parameter (below roughly 50%, and often below 25–30% for tighter control) alongside temperature, not left to general cleanroom defaults. ^[33]

Conclusion

Diamond turning earns its place in ophthalmic manufacturing because it is one of the only subtractive processes capable of leaving an optical-quality surface without polishing, and because it can cut geometries that molding and spin-casting cannot easily reproduce. But it is not a black box that guarantees nanometer finish on any material placed on the spindle. Surface finish is set by a feed-and-tool-radius geometry that behaves as a best case, degraded in practice by tool wear, machine vibration, thermal drift, and material-specific behavior — and ophthalmic materials introduce their own specific complications on top of that general picture: hydrogel materials that are cut in one physical state and worn in another, RGP polymers whose chemistry has been tuned around machinability and heat resistance as much as around biocompatibility, and IOL materials where the real precision-manufacturing question is often not "how well can this be diamond-turned" but "should this be diamond-turned directly, or diamond-turned once as a mold and then replicated thousands of times over." Getting that surface right is not a finishing-department detail — in a contact lens or an IOL, the machined surface is in the optical path of a patient's vision, and the roughness left behind is measured in the same nanometer units that determine whether light focuses cleanly or scatters.

References

1. The investigation of influence of tool wear on ductile to brittle transition in single point diamond turning of silicon. ScienceDirect. Available at: https://www.sciencedirect.com/science/article/pii/S004316481630165X

2. Brittle Machining Issues in Single Point Diamond Turning of Aspheric Infra-Red Optics. Available at: https://www.academia.edu/85488154/Brittle_Machining_Issues_in_Single_Point_Diamond_Turning_of_Aspheric_Infra_Red_Optics

3. The investigation of influence of tool wear on ductile to brittle transition in single point diamond turning of silicon (tool-wear-triggered transition). ScienceDirect. Available at: https://www.sciencedirect.com/science/article/pii/S004316481630165X

4. Review of single-point diamond turning process in terms of ultra-precision optical surface roughness. ResearchGate. Available at: https://www.researchgate.net/publication/338000416_Review_of_single-point_diamond_turning_process_in_terms_of_ultra-precision_optical_surface_roughness

5. Ductile behaviour in single-point diamond-turning of single-crystal silicon. ResearchGate. Available at: https://www.researchgate.net/publication/232402224_Ductile_behaviour_in_single-point_diamond-turning_of_single-crystal_silicon

6. Study on the control of surface roughness in single point diamond turning. ResearchGate. Available at: https://www.researchgate.net/publication/271483951_Study_on_the_control_of_surface_roughness_in_single_point_diamond_turning

9. Influencing Factors and Theoretical Models for the Surface Topography in Diamond Turning Process: A Review. PMC. Available at: https://pmc.ncbi.nlm.nih.gov/articles/PMC6562771/

10. Origins for the size effect of surface roughness in diamond turning. ScienceDirect. Available at: https://www.sciencedirect.com/science/article/abs/pii/S0890695516300293

11. Signal analysis of surface roughness in diamond turning of lens molds. ScienceDirect. Available at: https://www.sciencedirect.com/science/article/abs/pii/S0890695504001452

12. The effect of feed rate and of tool nose radius on the roughness of oblique finish turned surfaces. ScienceDirect. Available at: https://www.sciencedirect.com/science/article/abs/pii/0043164873900197

13. A review of tool wear mechanism and suppression method in diamond turning of ferrous materials. Int. J. Adv. Manuf. Technol., Springer. Available at: https://link.springer.com/article/10.1007/s00170-021-06700-8

14. A critical review on the chemical wear and wear suppression of diamond tools in diamond cutting of ferrous metals. IOPscience. Available at: https://iopscience.iop.org/article/10.1088/2631-7990/ab5d8f

15. White Paper on Ultra-Precision Turning. Dr. Thomas Liebrich, RhySearch, November 2024. Available at: https://www.rhysearch.ch/fileadmin/user_upload/Dokumente/whitepaper/UltraPrecisionTurning_eng_Liebrich.pdf

16. Machinability of hydrophobic acrylic polymer for customized intraocular lenses. ResearchGate. Available at: https://www.researchgate.net/publication/396119683_Machinability_of_hydrophobic_acrylic_polymer_for_customized_intraocular_lenses

17. On the precision single-point diamond machining of polymeric materials. ScienceDirect. Available at: https://www.sciencedirect.com/science/article/abs/pii/S0924013606001658

19. The manufacture of contact lenses. Optician Online CPD archive. Available at: https://www.opticianonline.net/cpd-archive/6066

20. ISO 13485 for Ophthalmic Device Manufacturers: The Complete Compliance Guide. Rotlex. Available at: https://rotlex.com/iso-13485/

21. High precision rigid gas-permeable contact lens edge curve polishing lathe. U.S. Patent 9,434,042. Available at: https://image-ppubs.uspto.gov/dirsearch-public/print/downloadPdf/9434042

22. Optical Scattering and Surface Roughness. Eckhardt Optics. Available at: https://www.eckop.com/resources/scatterometer-resources/optical-scattering-versus-surface-roughness/

23. Process for hydrating soft contact lenses. U.S. Patent 5,080,839. Available at: https://image-ppubs.uspto.gov/dirsearch-public/print/downloadPdf/5080839

24. Method for lathing a lens. U.S. Patent 6,315,650. Available at: https://image-ppubs.uspto.gov/dirsearch-public/print/downloadPdf/6315650

25. Diamond Turning of Polymer Optics. NC State Precision Engineering Consortium. Available at: https://pec.ncsu.edu/research/diamond-turning-of-polymer-optics/

26. Derive Formula for Surface Roughness in Turning with a Rounded Tool. Available at: http://www.minaprem.com/machining/principle/quality/derive-formula-for-surface-roughness-in-turning-with-a-rounded-tool/

27. Chapter 6: RGP Materials. In Contact, Contamac. Available at: https://contamac.com/wp-content/uploads/2024/12/In-Contact_Chapter-6.pdf

28. In-vitro glistening formation in six different foldable hydrophobic intraocular lenses. PMC. Available at: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7938589/

29. Evaluation of in vitro glistening formation in hydrophobic acrylic intraocular lenses. PMC. Available at: https://pmc.ncbi.nlm.nih.gov/articles/PMC3732199/

30. ISO 13485 for Ophthalmic Device Manufacturers: The Complete Compliance Guide (biocompatibility/ISO 10993 context). Rotlex. Available at: https://rotlex.com/iso-13485/

32. Manufacturing a Brand-New Industry. Review of Contact Lenses. Available at: https://www.reviewofcontactlenses.com/article/manufacturing-a-brandnew-industry

31. Generic Vibration Criteria for Vibration-Sensitive Equipment. Colin G. Gordon, Colin Gordon Associates (published SPIE 1991 / IEST 1993). Available at: https://www.octava.info/files/%D0%A1%D1%82%D0%B0%D1%82%D1%8C%D0%B8/Gordon-SPIE99.pdf

33. Method for lathing silicone hydrogel lenses. U.S. Patent 9,248,614. Available at: https://image-ppubs.uspto.gov/dirsearch-public/print/downloadPdf/9248614

34. Contact lens having improved dimensional stability. U.S. Patent 6,265,465. Available at: https://image-ppubs.uspto.gov/dirsearch-public/print/downloadPdf/6265465

Note on sources: all figures, tolerances, and findings above are drawn directly from the indexed technical and peer-reviewed sources listed. Where a claim is standard machining theory (e.g., the geometric roughness relationship) rather than a single study's finding, this is noted in context. No data points were estimated or extrapolated beyond what the cited sources report.