The Inside (Technical) Scoop on Lens Coating Durability

[Nodda Duma]

I was going to respond to one of the various currently active threads on lens cleaning, but thought I might start a separate post. I thought it might be informative to provide "the inside scoop" -- i.e. the technical info that I need to know when design and specifying optics and optical coatings for durability requirements (a.k.a. abrasion resistance). From a historical context, it will also give you insight into why older single-layer and multi-layer coatings are softer than uncoated glass and newer multi-layer coatings. Furthermore, if you dig into the MIL-PRF-13830B standard that I reference below, you will also gain some insight into the type of "cleaning abuse" different coatings will survive (material and force applied). The book I reference, Thin Films for Optical Systems, is an industry standard text -- by that I mean every coating designer (and lens designer worth his/her salt) has a copy of this book on their shelf...likely with plenty of bookmarks and coffee stains scattered throughout, and much of the information used in their daily work. I'm posting this in the 35mm sub-forum, but it applies to all optics used in the consumer market.

Here is the information on anti-reflection (AR) coating durability.

Within the industry, the "toughness" of an optical material is commonly specified using Knoop Hardness (you can look this up to see what it means). There are other metrics, but this is the one I most commonly see used in the design industry and is what I use when working with the opto-mechanical engineer. Resistance to scratching is tested using "cheesecloth" and "pencil eraser" tests, a.k.a. Moderate and Severe Abrasion tests described in MIL-PRF-13830B Appendix C (if I remember correctly). Expect that single-layer coatings (MgF2) and older multi-layer coatings will pass the moderate abrasion test but fail the severe abrasion test, and that modern multi-layer coatings will pass the Severe Abrasion test. There is a caveat to this last statement ... use cost and reputation of the manufacturer as a guide. We can discuss the historical context of when different coatings became commonplace in follow-up posts... I don't want to detract too much from this initial post. Note that this abrasion resistance test is only one of several durability callouts: Humidity and adhesion requirements are also commonly called out and tested separately...the information is contained in the reference standard mentioned above.

I'll provide Knoop Hardness values for materials used in the coatings, and then compare them to representatives of the different groups of optical glasses used for designing camera objectives.

Anti-reflection coatings typically use the following materials: MgF2 (magnesium flouride), TiO2 (Titanium Oxide), SiO2 (Silicon Oxide), and Al2O3 (Aluminum Oxide). As coating specifications become more complex, other materials such as ZrO2 (Zirconium Oxide) and Ta2O2 (Tantalum Oxide) are used as coating layers to meet the requirements. For example, I would expect to see combinations of these materials used in the coatings for dielectric filters with sharp cutoffs such as narrow bandpass filters or notch filters, or extremely wide spectral bands (VIS / NIR) and/or wide temperature requirements (i.e. military standard temperature ranges). Thorium Oxide used to be used as well (its use for coatings lasted longer than its use as an optical material), but is no longer used for obvious reasons.

The common materials mentioned above are typically used in the following manner:

Single Layer AR Coatings are MgF2, with a Knoop Hardness of ~415. It is slightly soluble in water (I've seen the damage from condensation myself under microscope) and minimizes reflections at only one wavelength. Typically that wavelength is green -- at the center of the visible band -- which explains the blue or violet cast of reflections off single-layer coatings.

Common (almost exclusively used in the consumer market) Multi-Layer AR coating Materials are TiO2 - SiO2, either as a two-layer coating, or multi-layer combinations of these materials.

The Knoop Hardness of the coating varies with the combination of the materials, as well as the deposition method (how it is deposited). Here are some examples:

TiO2 - SiO2, evaporation (older method): HK = 410
TiO2 - SiO2, ion deposition (newer method): HK = 730
TiO2 - SiO2 plus SiO2 outer layer, ion deposition: HK = 1100
ZrO2 - SiO2, ion deposition: HK = 940
Ta2O2 - SiO2, ion deposition: HK = 970

You can compare this to Knoop Hardness of different optical glasses by reviewing the data in the Schott Optical Glass Datasheet linked below. The value to look for on each glass data sheet is HK in the "Other Properties" table. Consider these representative of the range of hardness values of the optical glasses used in the design of camera lenses. The page #'s in the table of contents are linkable:

http://www.schott.com/d/advanced_op...ss-collection-datasheets-english-17012017.pdf

N-BK7 is a typical crown. (HK = 610)
F2 is a typical flint. (HK = 420)
N-LAK9 is a typical lanthanum glass (HK = 700)

Quick side note: N-LAK9 is the first lanthanum glass, which was developed for and used by Leitz in the Summicron design of the 1950s. It and other lanthanums are very commonly used in fast lenses today.

Here are Knoop Hardness values for substrates commonly used for filters / windows:

Borosilicate (common filter / window material) has a Knoop Hardness of 480.
Fused Silica (more expensive filter / window material) has a Knoop Hardness of ~820.
Sapphire -- a common window material for non-consumer applications -- has a Knoop Hardness of 1800 - 2200 depending on which crystal lattice axis the stress is acting upon.

References:
Thin Films for Optical Systems by Flory, pg 470
Antireflective Coatings: Conventional Stacking Layers and Ultrathin Plasmonic Metasurfaces, A Mini-Review by Hedayati and Elbahri
Schott Optical Glass Datasheets as linked above

-Jason

 

[tedr1]

Thank you for providing this information.

Is there an easy way to determine the coating type by visual inspection of the colors of the reflections? I note you give an example above for single coating, do the other coatings have reliable colors too?

 

[Nodda Duma]

Thank you for providing this information.

Is there an easy way to determine the coating type by visual inspection of the colors of the reflections? I note you give an example above for single coating, do the other coatings have reliable colors too?

You're very welcome! And, unfortunately, no.

While you can readily determine by observation which surfaces are coated, it is almost impossible to determine exactly what the materials are (beyond the knowledge that single-layer AR coatings in the visible spectrum are exclusively MgF2). That said, with experience in coating specifications, the end-use for the lens in question, and familiarity with which materials and coating methods are/were commonly used at the time the lens was manufactured, you can make a pretty good guess at what the coating layers *may* look like.

I recently ran across a journal paper which discussed a method to inspect a multi-layer coating in a deterministic manner. While I don't recall the specific method, the means involved are far beyond the reach of the individual as well as most optical labs and I don't believe the lens survived the inspection.

Interestingly, coating design is a very specialized art which, like RF design for Electrical Engineering, borders on black magic (even more so than for my own profession of lens design). If you are a talented, experienced coating designer, you are pretty much guaranteed a lifetime of 6-figure employment anywhere in the world that an optical shop is located. How you get that experience usually involves a fair amount of "right place at the right time" to learn from a worthy mentor or having the opportunity to learn on-the-job.

 

[Sirius Glass]

Good information BUT ... which coating and hardness do I have it I have, for example, the following lenses:

• Zeiss T lens
• Zeiss T* lens
• 1930's lens
• 1940's lens
• 1950's lens
  
• Nikon, Minolta, Canon, Leitz, ... 1960's lens
• Nikon, Minolta, Canon, Leitz, ... 1970's lens
• Nikon, Minolta, Canon, Leitz, ... 1980's lens
• Nikon, Minolta, Canon, Leitz, ... 1990's lens
• Nikon, Minolta, Canon, Leitz, ... 2000's lens

 

[Nodda Duma]

With an understanding of what coating technologies were introduced when, I would expect --

Good information BUT ... which coating and hardness do I have it I have, for example, the following lenses:

• Zeiss T lens -- single coat MgF2
  
• Zeiss T* lens -- Depends on the manufacture date of the lens. T* = two-layer (early) / multi-layer (later)
  
• 1930's lens -- uncoated (oxidized)
  
• 1940's lens -- single coat MgF2 (if coated)
  
• 1950's lens -- single coat MgF2 (if coated), or two-layer evaporation
  
• Nikon, Minolta, Canon, Leitz, ... 1960's lens -- single layer or two-layer evaporation (use cost/markings as a guide)
  
• Nikon, Minolta, Canon, Leitz, ... 1970's lens -- single layer or two / multi - layer evaporation (use cost/markings as a guide)
  
• Nikon, Minolta, Canon, Leitz, ... 1980's lens -- single layer or two / multi - layer evaporation or ion deposition (use cost/markings as a guide)
  
• Nikon, Minolta, Canon, Leitz, ... 1990's lens -- single layer or multi - layer evaporation or ion deposition (use cost / markings as a guide)
  
• Nikon, Minolta, Canon, Leitz, ... 2000's lens -- single layer or multi - layer ion deposition (multi-layer is common by this time, use cost / markings as a guide)

Use that as a rough guide .. I'm sure the marketing departments would have touted each new technical advancement in the ads of the period, so it is entirely possible to determine exactly when new coatings were introduced for whom (and I do recall seeing some of that info in the APUG archives). However, that's a bit more digging than I have time for.

Here is a link to the history of optical coatings prior to 1960:

Dead Link Removed

And information on Zeiss T* coatings (there is some history in there)

http://lenspire.zeiss.com/en/wp-con...reduction-of-reflections-of-camera-lenses.pdf

 

[Nodda Duma]

By the way, here is the link to "Century of Optics" pdf document that OSA put together for its 100th anniversary. This is an all-encompassing write-up that summarizes advancements across the entire optical industry. It's a 100MB file, so be forewarned.

http://www.osa.org/en-us/100/bookshelf/centennial_book_table_of_contents/

 

[albada]

Thanks for posting this information.

I didn't know that any coating materials were water soluble, and that might explain why Xenon 50/2 lenses found in Retina IIa cameras (early 1950s) often have haze on the coatings of the two surfaces that face the shutter blades. Those two surfaces are exposed to ambient moisture, and with poor air-circulation in there, moisture will stay around, damaging coatings.

Two questions: (1) Which materials (aside from MgF2) are water soluble? and (2) would it be possible to remove such a coating by keeping the surface wet for a few hours? I would like to remove the haze on those Retina lenses, and I wouldn't mind removing the coating on those two surfaces.

Thanks,

Mark Overton

 

[Nodda Duma]

Hi Mark,

None in the materials listed. Let me clarify that MgF2 is *only very slightly* soluble in ph-balanced water. It is worse when acidic or alkaline but not critically so. In most cases it will not be an issue unless you put it away with condensation left on the surface after shooting near the ocean or after acid rain. Let the surface dry out, or keep dessicant with the lens.

You could theoretically strip the coating off, but it is difficult to do properly outside of an optical shop (even then it's tricky). You strip it off in the same way you figure a surface -- with pitch and polish -- so you risk changing the surface figure and thus the prescription. I was tempted to try once but ended up not having to. If you are able to strip it, you either lose contrast or introduce the risk of ghost reflections (for single layer) or it's impossible to do so -- even for an optical shop -- if it is the outer layer of a multilayer coating. In a multilayer coating, each layer is a contributer to the AR performance, so you risk making reflections even worse than bare glass. Doesn't hurt to try, though, but easier to just replace the lens with an undamaged one.

Of the materials listed for coating use, none others are soluble. There are other optical materials used as optical elements which are soluble, but those limitations are accommodated in the design (i.e. hermetically sealed lenses). Few to worry about in the vis, most are exotic infrared materials. One interesting example is the material which was commonly used for windows in older infrared spectrophotometers: Sodium Chloride. Those windows were critical elements and required special handling.

Some optical glass families have relatively low acid or chemical resistance, but nothing remarkably drastic. Good designers will not use those materials for external elements and the surfaces are nowadays protected by coatings anyhow.

 

[E. von Hoegh]

I have an old 6" Goerz NY Dagor, ca 1908-10. The front and rear external surfaces have the most gorgeous bloom I've ever seen, a rich purply blue. Has there been any progress in the artificial aging of surfaces?

 

[Nodda Duma]

I have an old 6" Goerz NY Dagor, ca 1908-10. The front and rear external surfaces have the most gorgeous bloom I've ever seen, a rich purply blue. Has there been any progress in the artificial aging of surfaces?

That "tarnishing" or oxidation was the inspiration for the development of AR coatings, so in a way .. yes ! AR coatings combined the advantages of that blooming / tarnish (whatever you want to call it) with repeatable and consistent performance in manufacture.

Optical systems for critical applications (i.e. not cheap crap) actually do experience accelerated aging via Qualification and/or Validation/Verification testing which are SOP for modern engineering design & development (for prototypes, initial production examples, and samples pulled from production lots). MIL-STD-810G provides industry-wide guidance for environmental testing, and design requirements typically incorporate a subset of that standard to verify the design meets environmental requirements. Temp, humidity, vibration, temp shock, etc. etc. .. Essentially the crap is beaten out of the product to ensure it still works. That's common practice for anything designed by engineers (including optical systems).

 

[bernard_L]

Dead Link Removed

Thank you a lot for this link. Where I learned that the origins of Lippmann plates are in a book by Goethe in 1810...

 

[E. von Hoegh]

That "tarnishing" or oxidation was the inspiration for the development of AR coatings, so in a way .. yes ! AR coatings combined the advantages of that blooming / tarnish (whatever you want to call it) with repeatable and consistent performance in manufacture.

Optical systems for critical applications (i.e. not cheap crap) actually do experience accelerated aging via Qualification and/or Validation/Verification testing which are SOP for modern engineering design & development (for prototypes, initial production examples, and samples pulled from production lots). MIL-STD-810G provides industry-wide guidance for environmental testing, and design requirements typically incorporate a subset of that standard to verify the design meets environmental requirements. Temp, humidity, vibration, temp shock, etc. etc. .. Essentially the crap is beaten out of the product to ensure it still works. That's common practice for anything designed by engineers (including optical systems).

Thanks. I remember from reading/research that in the 1890s H.D.Taylor and some others were looking for a way to produce this bloom reliably, nobody really had any success until the late 1920s, with the first modern coatings appearing in the early/mid 30s.

I just wish that bloom was on the inner surfaces of my lens.

I've also noticed that on some old (80 - 125 yrs) lenses the surface presents a strange almost dark reflection, placed next to other uncoated lenses these lenses have less brilliant reflections, but no pretty colors. Looking through the lens, all appears normal. I assume this is a similar tarnishing, just of the wrong thickness to produce colorful reflections?

Thanks again for this thread.

 

[albada]

Thanks for the info. Yes, I was afraid of that. Removing coating is all but impossible, and would require specialised grinding equipment, and having that I might as well make my own lenses.

Mark Overton

Hi Mark,

None in the materials listed. Let me clarify that MgF2 is *only very slightly* soluble in ph-balanced water. It is worse when acidic or alkaline but not critically so. In most cases it will not be an issue unless you put it away with condensation left on the surface after shooting near the ocean or after acid rain. Let the surface dry out, or keep dessicant with the lens.

You could theoretically strip the coating off, but it is difficult to do properly outside of an optical shop (even then it's tricky). You strip it off in the same way you figure a surface -- with pitch and polish -- so you risk changing the surface figure and thus the prescription. I was tempted to try once but ended up not having to. If you are able to strip it, you either lose contrast or introduce the risk of ghost reflections (for single layer) or it's impossible to do so -- even for an optical shop -- if it is the outer layer of a multilayer coating. In a multilayer coating, each layer is a contributer to the AR performance, so you risk making reflections even worse than bare glass. Doesn't hurt to try, though, but easier to just replace the lens with an undamaged one.

Of the materials listed for coating use, none others are soluble. There are other optical materials used as optical elements which are soluble, but those limitations are accommodated in the design (i.e. hermetically sealed lenses). Few to worry about in the vis, most are exotic infrared materials. One interesting example is the material which was commonly used for windows in older infrared spectrophotometers: Sodium Chloride. Those windows were critical elements and required special handling.

Some optical glass families have relatively low acid or chemical resistance, but nothing remarkably drastic. Good designers will not use those materials for external elements and the surfaces are nowadays protected by coatings anyhow.

 

[Nodda Duma]

Thanks for the info. Yes, I was afraid of that. Removing coating is all but impossible, and would require specialised grinding equipment, and having that I might as well make my own lenses.

Mark Overton

If you ever *do* set up to grind lenses, let me know and I'll feed you some designs.

-Jason

 

[Patrick Robert James]

Thanks for posting this Jason. I've appreciated your knowledge of the subject in the past but haven't properly acknowledged it. Your expertise is much appreciated.

If you ever get set up to coat lenses I would love to get the uncoated 135mm f/3.5 Eurynar I have coated. Not hoping for much, but let us know! Lol.

As an aside, I have found the EBC coating Fuji put on their large format lenses to be fascinating. The elements are almost invisible, which to me means the coating really works! That is why I use them. Great lenses.

I hope you continue to post whatever interests you here. I enjoy reading it.

 

[Nodda Duma]

It is my sincere pleasure to add to the shared knowledge of the forum in exchange for all the information I've put to use.

 

[Sirius Glass]

This thread should be a sticky or in the resource section.

 

[RalphLambrecht]

Good information BUT ... which coating and hardness do I have it I have, for example, the following lenses:

• Zeiss T lens
• Zeiss T* lens
• 1930's lens
• 1940's lens
• 1950's lens
  
• Nikon, Minolta, Canon, Leitz, ... 1960's lens
• Nikon, Minolta, Canon, Leitz, ... 1970's lens
• Nikon, Minolta, Canon, Leitz, ... 1980's lens
• Nikon, Minolta, Canon, Leitz, ... 1990's lens
• Nikon, Minolta, Canon, Leitz, ... 2000's lens

great info; thanks for that Nodda

 

[flavio81]

Bumping up this thread Jason. Very interesting info.

I am very curious about the usage of two layer coatings in the 60s, besides Minolta's "achromatic" coating. I had a Canon lens FL 85-300/5.0 which was marketed in 1964, and looking at the glass reflections there were many colors which made me think there was some sort of dual layers there. This lens has 9 groups.

I also have a FL 19/3.5 of 11 elements (1965) and (i think) 8 or 9 groups, and it apparently, might have dual layer. Otherwise i don't explain how they had dared to build such a lens.

Both those lenses were pretty expensive in their day, the 85-300 was canon's most expensive lens.