Originally Posted by
cyccommute
How much solute dissolves in a solvent is dependent on the identity of both and the interactions between the solute and solvent. Some solvents can dissolve infinite amounts of solutes like sugar in water, ethanol in water, acetone in water or even oil in mineral spirits. Other solutes only have limited solubility in certain solvents. Even the measures of solvent "strength" depend on the solute and the solvent as well as what you are measuring.
And that doesn't even take into account the cost, toxicity or flammability of the solvent. d-Limonene may be a great solvent but it costs around $40 a gallon. A gallon of mineral spirits cost $15 per gallon.
MEK and xylene cost around $20 each but are more flammable and more toxic. The reason I list xylene rather then toluene is that toluene is more difficult to find because of its toxicity. For the casual home user, mineral spirits is a better choice for cleaning chains and and bearing surfaces. It's cheap, relatively safe and efficient.
In my experience, mineral spirits will take care of even varnish deposits in 5 to 10 minutes.
Honestly, 20 microns or 0.02mm is good enough for a bicycle chain. One micron...0.001mm...is overkill for the application. And it won't remove the dissolved oils at all.
It is possible to over think this and trying to find a "better" solvent or a filter to take out smaller particles is certainly overthinking the problem.
Yep, I'm aware of miscibility and we are speaking of non-polar solutes.(except dumondtech)
toluene and xylene are very similar in most respects, plain benzene is by far the toxic one of the family. I didn't list it as the first choice(mostly due to price).
d-limonene is actually fairly toxic but it makes people think "natural and fruity". It also has emulsifying properties that petrol distillates don't, and it works well in low concentrations, the cost is a small factor if it is reused many times.
I haven't actually seen a bicycle part develop varnish, I have meet varnishes in other mechanics that laughed at a week long soak in kerosene, but took only a few hours in a warm bucket of nitroamine solution("carb dip" more of a chemical reactant than a solvent, water rinse-able end product); and the grease cake(no oil left in it and is found in vintage bicycles) that resisted gasoline and kerosene for days with daily shaking, but in one day with toluene/MEK soaking it slipped right off (destroyed my nitrile gloves in a couple of minutes too)
With the right containers[w/lids too] it only takes a few ounces of the strong solvents, once dissolved the final rinse is with the cheap and easy stuff. I've been working with the same few quart cans for a decade, not because of filtering but just good thrifty use and getting a few uses out of each batch.
As for filters I was presenting options, not suggestions. Which reminds me; if oil becomes excessively concentrated in the solvent they can be separated with simple distillation as is used with commercial parts washers, this would also leave behind any hard abrasive contamination.
While it isn't worth the effort for a bike chain due to rapid re-contamination with road dust, a sub micron filtration of the lubricant would [if kept clean] significantly extend the wear life of the parts as they don't operate within the hydrodynamic or elastohydrodynamic lubrication regimes, due to the slow speeds and start-stop oscillation type movement, the chain is in a boundary-heavy mixed state most of the time, meaning that there is less than a micron of lube and lots of direct metal contact between the microscopic high points of the surfaces while under load.
I usually end up tossing my well used solvents on the fire.