RE: [AV Media Matters] The Permeability of Polycarbonate (PC) use d in optical

[e.ha.zwaneveld@nfb.ca]

Greetings,

Joe Wrobel wrote:
For those of you who would like to learn more about the diffusion of
water into polycarbonate, I suggest you obtain a copy of the paper I
wrote that discusses the subject in depth:  "Ramp profiles for optical
disc incubation", SPIE Vol. 2338 Optical Data Storage, 191-202, 1994.
Ed, you should be aware of this paper as it is referenced in the ANSI
standard for estimating life expectancy of CD-ROM (ANSI/NAPM
IT9.21-1996) and that for CD-R (ANSI/PIMA IT9.27-1999).

The purpose of the paper was to address the changes in polycarbonate
water content that occur as conditions are cycled during the high
temperature, high humidity incubations that are performed in accelerated
keeping studies of optical media using polycarbonate substrates.  As Ed
points out, the rate of water diffusion in and out of the polycarbonate
substrate is very different at its two surfaces.  Water diffusion
through the metal coated side is essentially nil; water diffusion
through the free polycarbonate surface, though slow, is appreciable over
the duration of an incubation experiment, especially those run at high
temperature.

The problem is not the moisture intake.  The saturated moisture content
of polycarbonate at 20C is 0.3%; at 80C it rises only to 0.5%.  The
problem is that unless due precautions are taken, when a disc held at
high humidity and high temperature for a long time is introduced into a
room environment, the moisture preferentially diffuses out the free
surface and, as it does, leaves behind a non-uniform distribution of
water in the substrate.  Since polycarbonate expands when it picks up
water, the gradient in water content causes the polycarbonate substrate
to symmetrically deform ("bowl") to a spherical surface to accommodate
the gradient in expansion.  If the gradient is extreme, the disc
deflection could exceed the CD deflection specification until the disc
equilibrates with the environment.  And at room temperature, this would
take a rather long time, up to several days.

The latter point is relevant to the discussion of whether a brief (10
minute) immersion of a CD in water will do any harm.  The answer is no.
The diffusion constant, D, of water in polycarbonate is 7.5*10^-8 cm^2/s
at 25C.  The depth of water diffusion over time t is roughly equal to
the square root of the product of D and t.  In ten minutes, the depth of
moisture diffusion in polycarbonate at room temperature is calculated to
be 0.07 mm.  In that depth, the moisture content will rise from its
equilibrium value (0.15% at 25C, 50%RH) to its saturated level (0.3% at
25C).  And, if the disc is dried, approximately ten minutes later the
moisture equilibrium in the polycarbonate will be reestablished.

Reply:
Thank you Joe for a most helpful and well-supported argument. Your calculation would indicate that 10 minutes of water exposure will do no harm to the disc. Graeme Jay, referring to one system used for polishing points out that with one system, the amount of water is measured in drops. But in another discussion, it was apparent that the grinding procedure can take much longer and on various disc polishing homepages we are told that the process can be repeated multiple times. And if I am not mistaken, some of the industrial systems reported, use a lot more water than a few drops. So we may concluse that polycarbonate water permeability may be a non-issue.

Joe stated:
Interestingly enough, since pure DVD discs are symmetric, they require no humidity barrier layer and have none.  Moisture comes and goes depending on the environment, but because any resultant gradients are symmetric about the center of the disc, no deflection results.

That sounds logical enough. I would like to draw your attention to a test report that relates to various types of ink used on CD and DVD media, and how they cause unwanted deflection, authored by Guy Masse, of Encres Dubuit, entitled: "Printing with DVD inks: looking beyond shrinkage", in One to One, May 2000. In the report, tests are described with various CD printing inks as they were tested on DVD media. After finding that most inks were unsuitable for various reasons, one potentially acceptable ink was applied that had good polycarbonate adhesion and low shrinkage after UV drying. The results of both tangential and vertical deviation and before the climate test, was nice and flat and within tolerance. Then a typical climate test in the form of exposure to 80 degrees C, and a relative humidity of 85% for 72 hours was applied. The results were disastrous, making the disc look like a Chinese hat. This serious degree of deviation was determined to have been caused by humidity intake during the climate test. Some formulations of ink can apparently absorb water out of the steam. They cause the ink to swell and that in turn deforms the disc. And that was not just a temporary effect. The logical conclusion drawn was, that a different ink formulation is required for DVD printing than for CD printing. It must have low shrinkage and must not absorb humidity. Note that this occurs in spite of the fact that the disc halves of DVD are theoretically in balance and warping or cupping or Chinese hatting, should not occur. If the ink formulation alone can cause a DVD disc to take on the form of a Chinese hat, who can tell what kind of ink was used on a DVD or SACD disc presented for polishing? We have in addition reports, such as from Ed Mino at TDK, that paper labels also cause warping and render discs unplayable.

In addition, Joe provided an insight and reflections relating to the SACD application. He reminded us of the effort invested by Kodak to test the media it markets and pointed to a website to be better informed. That is highly commendable. But the data shown there is five years old and a lot happens in five years in the optical disc business, with falling prices, new suppliers, conflicting vendor claims and various alternatives to dyes, their filters and stabilizers and reflective layers. Is that the most recent time Kodak tested the quality of the disc media it sells? We must appreciate that a large percentage of CD media are not made to the kinds of specifications Kodak respects. The independent Testing Laboratory, Media Sciences, Inc. (see www.mscience.dom) reported recently that CD-R disc quality from experienced manufacturers has 'improved' from 30% defective in 1998 to 'only' 13% defective in 2000. Failures for other discs increased from 33% in 1998 to an alarming 60% in 2000, mostly for radial tracking and jitter.

Joe Wrobel then asked:
With regards to SACD, could you please direct me to the data that suggests it will have a long life?  In fact, could you direct me to any study on the estimated lifetime of any DVD format?

And in the posting that he responded to I stated: "I highly recommend the evaluation of the SACD systems that are now available and thorough validation testing of the benefits of the humidity barrier (s) included in SACD media." In fact that advise relates to all recording media, including DVD and CD media.

Users with serious disc applications are not well-served by the unwillingness of most manufacturers to test,publish and update information that enables users to make an intelligent choice, rather than be forced to act on good faith.

In thus far unpublished and the most recent manufacturers data, the life expectancy of CD/DVD-ROM, CD/DVD-R and DVD-RAM, is believed to extend beyond 50 years and possibly reaching 100 years. Phase-change media archival life projections at a range of temperatures and based on Arrhenius plots of accelerated life test data comparing three manufacturers, referred to as A, B and C, would seem to indicate that.

Thank you Graeme Jay and Joe Wrobel for your efforts in shedding some further (non-fading) light on the subject!

With best regards,

Ed H. Zwaneveld,
Technological Research and Development,
National Film Board of Canada, and
Chair AMIA Preservation Committee,
August 28,2000