Polarized light
is a pretty useful tool. Not only the two polarizers make a nice effect of going
dark/transparent depending on their relative angle, they can be also used for many actually
useful things, e.g. visualisation of strain in transparent plastics and glass.
New polarizer sheets are either small and expensive, or larger and VERY expensive. But each LCD panel usually has two, in the form of a glued-on plastic foil, one on each side of the glass sandwich. It's just about sourcing a cracked or otherwise useless display panel of a suitable size and prying the sheets off.
One LCD panel will yield two polarizers. Depending on the design, both can be fully transparent, or one of them, the front one, can have matted surface (for non-glossy displays). The bottom one may or may not be matted, but in the tested case it was found to be glossy and transparent.
A hat tip goes to Jeri Ellsworth, who shown the rudimentary process on YouTube
,
which provoked me to document the process in more detail.
The adhesives and sealants used in electronics often become softer at higher temperatures. This is exploited
for example for salvaging ferrite cores and bobbins
from discarded power supplies.
The display panel is removed from the backlight plate (which is useful too). Note the electronics on a flexible substrate, with directly bonded long narrow silicon chips; these are the row and column panel drivers, sending signals to the pixels themselves. In some cases these become gradually unbonded and cause failures involving a band of bad colors or failed pixels affecting an entire row or column. The isolated display is a thin layer of liquid (the liquid crystals themselves, usually a cholesterol ester but can be also another suitable organic chemical), on each side with a thin glass panel (about 1mm thick, fairly fragile glass). The sandwich's outer sides are laminated with plastic polarization filters.
 LCD panel, front side |  LCD panel, back side |  LCD panel, back, electronics exposed |  LCD panel electronics, column driver chips |
 Cracked panel, backlit |  Cracked panel, backlit |
The glass-and-plastic sandwich is placed next to the nozzle of a hot air generator (like the Hot Air Blower
or a hairdryer. Beware of too high temperatures, the plastic itself should not soften nor undergo any other
damage. Allow enough time for the heat to soak through the plastic and warm the glass and the glue layer between
the plastic and the glass. Avoid direct touch of the panel and the nozzle or other excessively hot material; the plastic
could warp or discolor or partially melt or lose the polarization properties or undergo other kind of damage.
Gently pry the plastic from the glass. Use a blunt tool. A table knife or butter knife
will do a good job. Be gentle and patient, push the blade between the glass and the plastic in parallel, avoid prying -
you'd risk damage of either the foil or the glass. Avoid pulling on the foil itself, however tempting it is
to just tear the foil off the glass; the adhesive is fairly good and you'd risk either cracking the glass or forming
a sharp kink on the foil and damaging it. The foils don't cope well with sharp bends. The adhesive tends to stick to the
blade; it's annoying but can be lived with. It can be alleviated with a bit of a suitable solvent (more later).
 Ungluing the sheet |  Ungluing the sheet |  Ungluing the sheet |  Unglued sheet, minor heat damage visible on the edge |
Once the sheet is removed from the glass, the icky part begins. The sheet is still full of the
adhesive. Which attracts
dust and fibers and sticks to everything and wrecks the transparency. So the glue has to go. Preferably with a solvent.
The first solvent to test was toluene, on the basis of educated guess and availability (the workshop's supply of solvents was found to be grossly inadequate). The compatibility with the plastic was tested by putting a drop on its edge, waiting for several seconds, then wiping it off; no attack on the material was observed. Same was performed on the glue, where a somewhat satisfying interaction was observed.
The glue did not dissolve, but swelled and got somewhat less sticky and could be scraped off. The material is a "supersnot" - a gooey gelly stuff with absorbed solvent, like what a god of glues blows from his nose. Work on old newspapers or other disposable substrate, as it will get everywhere. (Beware on the tendency of the printing ink to dissolve in some solvents, you could inadvertently transfer some onto the table below. Oops.)
The performance of the solvent was less than hoped for, the adhesive was fairly resilient and toluene stinks. The next day
more solvents were bought under the guise of paint thinners - a light petroleum fraction and a xylene. (Various solvents and
their combinations are available for different paint formulas; they usually contain one or more of aliphatic and aromatic
hydrocarbons, aliphatic ketones and esters, usually listed in small print on the can.) Other solvents sourced were turpentine
and the citrus oil (a mixture of terpenes) based sticker remover in the form of spray. Ketones-containing solvents weren't tried
out of concern for the incompatibility with the plastic sheet.
All the solvents tested shown similar performance. The citrus oil sticker remover won on the basis of ease of dispensing (just spray it on) and least-bad smell; turpentine, with its distinct smell of pine trees, was the second choice. Light petroleum and xylene, while also possibilities, were decided against on the basis of stink.
The solvent was sprayed on the adhesive layer and given a few minutes to soak in. The thin layer of adhesive turned into a thicker layer of swelled snot. A piece of tissue was then used to wipe the snot off the surface; it does not go well, the goo tends to recontaminate the surfaces and there is a lot of it. The layer also tends to stick fairly well, and needs to be mechanically scrapped; a fingernail does a good job.
It is tempting to scrape the goo off with a knife. Avoid it if you can; even a dull dinner knife turned out to be hard enough to scratch the soft plastic. The scratches then will be very well visible in dark field due to absence of polarization effect. Stick with slow, patient wiping, use copious amount of solvent, and do the work in several passes; first get rid of most of the glue, then of most of the rest, and only then go for removal of all the residues and smudges. A scraper made from a suitably soft material (softer than the foil but hard enough to not deform much under the force needed to separate the snot from the substrate) could be very helpful.
The polarizer is often present as a thin coating on the plastic substrate, on the glue side. As such, it is fairly sensitive to scratches and other mechanical damage. Be very careful if you want to minimize introduction of such defects.
The work is messy and takes a long time. A rerun of something on TV is a good companion for the job. Awareness of the price of the equivalent new polarizer sheets off-the-shelf is also helpful to keep the morale up.
It is advisable to perform the final cleaning using an antistatic-enhanced cleaning solution/foam. The cleaned and solvent-wiped foils were found to get electrically charged and attract dirt and stick to things before such treatment.
 Washing off the adhesive |  Removed adhesive, "supersnot" |
The LCD screens have more useful components. In addition to standard electronics, there is the backlight panel, useful for e.g. light table or a large-area room light. The glass layers from the display itself are coated with a matrix of transparent electrodes, which may find their own use in some exotic projects. One of the glass panels also contains the RGB color filters deposited under the electrodes.
More pics, pics of the polarizers in action, a frame for the polarizers, a backlight panel.