Hi I need to find a method of detecting less than 1 ppm contaminants in water or ethanol.
TDS meters work for the most part, but are often too unwieldy for my purposes and rarely can detect sub 1ppm levels of contaminants.

Any ideas for other ways? pH or some other method?

Are you serious? GC/MS/IR FTW.


PH my ass.

would work but it is far too expensive, I have a 500 dollar cap for finding this out.

Send it to an analytical lab.

I need to do this quietly, its not illegal, but I need to build the prototype so I can get a patent.

So if I could get equipment to do this for under 500 bucks, it would be grand

what are the detecting? metals? organics?

my first thought is ion sensitive electrodes for metals, organics pose a large problem if you don't know what you are looking for. maybe concentrating down a known volume and running tests on that will give a more reliable result.

Are you serious? GC/MS/IR FTW.


This.

Okay, get a million litres of water/ethanol...

No but srsly, you're going to have to find some way of putting it through some sort of analytical machine. IR is possibly the cheapest while not compromising on quality and use, but a desktop model still costs about... $6000 I think.

Make friends with people in universities / colleges, or students well into degree courses, and convince them to run a sample for you.

Okay, get a million litres of water/ethanol...

^^ this.

On a... well, honestly, a less-serious note...

- does it have to be foolproof, or just a crude estimate/assay?

- does it have to identify the shit, or just provide a first-pass estimate on where the best place to start looking for, say, groundwater contaminants might be if you've got too many sites and not enough analytical budget?

...and last but not least...

- if you're looking for a specific contaminant, fucking well say so, lol.

Anyways... if the first two are, umm... "loose-tolerance," as it were... electroconductivity. It is your friend.

By virtue of water being a polar solvent, most things that dissolve in it will serve as some form of disassociation-inducing electrolyte. Several of these will be utter crap - I don't imagine the micromiscibility of diethylether is going to provide "1337 4w3s0m3 suuprkonduktivity lol"...

...but if you can detect nanoampres with obscene amounts of sensitivity, you can probably get a semi-decent estimate on whether the conductivity on, say, .01ml of an aqueous sample has the same conductivity as the impressive resistance of completely-deionized water or not.

But, as long as you don't mind a slight risk of missing a few bizzare compounds and other absences of absolute guarantee, and also don't give a fuck about identifying whatever the hell this 1ppm might be (for which you damn well get yourself a million fucking litres and condense it down)... electroresistance at a known electrode seperation is probably the most reasonable way to get remarkable sensitivity at reduced cost...

...plus, it makes a really kick-ass instrument, since you get some hand-held crap with... A PROBE!! That AUTOMATICALLY SPITS OUT A PASS/FAIL ANSWER!! lol...

You might still blow your $500 budget trying to detect the fractional-picoampre difference in some low-ionization contaminant @ 1ppm... but, keep the electrodes about 1mm apart to reduce base resistance, you should get an approximate something, and table salt should make even a microampre guage jump at low concentrations at low seperation...

Good luck, buddy.

i'm not sure IR can be used quantitatively. at least, ive never seen or heard of it.

trace organics really have to be done by GC (and mass spec if you are interested in what they are). metals are significantly easier, with AAS a nice cheap option, although still completely out of range of a normal person.

some things like chloride and sulfates can be determined pretty easily without (too much) equipment, but at 1 ppm (where are you getting that figure from) significant sample prep would be required.

anyway, without knowing what your wanting to measure, im just pissing in the wind.

" - if you're looking for a specific contaminant, fucking well say so, lol."

reminds me of someone who wrote " fuck the shut up"



And while Im not providing the quote my answer would be THAT along the same lines as THIS

for 1 ppm....pretty much nothing, gc, ir, hplc, specs are not that specific, you could distill, concentrate, then test but you would need at least a 10X concentration, 1ppm is pretty much nothing, and to be realistic to see 1ppm you would need a test with resolution/margin of error in the 1 pptenmillion range

good point

I'm needing to find the concentrations of drugs in urine if that helps, I can't give away any more info

does anyone know what the total mass of a 1 or 2 mg flunitrazepam pill?

I lol'd fucking HARD. Kid, you aren't going to do your own piss tests, you have a lot of learning to do. First I suggest you find out the kind of test you are trying to beat, and if they even test for that. This is some serious lulz. All these high aspirations and secrecy for a fucking drug test.


Also, I suggest you get to working on filling that million gallon jug with piss, and get to distilling.

Since I just got notice from my lawyer it is protected idea I can tell you what I've been making.

It in essence is a field test for finding the drug levels in cattle urine, if you've ever worked for a vet in Alberta for a few weeks you'd see how useful it can be.

but seriously what is the total pill mass?

I doubt anyone here could tell you the total mass of a pill without having it in front of them and weighing it. Cant you get a scale and a pill and just find out yourself?

And if the urine tests pos. Do you tip them over as a sentence. ( after getting the bulls name that sold the drug of course)

It in essence is a field test for finding the drug levels in cattle urine, if you've ever worked for a vet in Alberta for a few weeks you'd see how useful it can be.

Well... that whole "urine" thing throws off the "pray it's a mild electrolyte" schtick...

Generally speaking, the only "field test" with that kind of sensitivity in a complex mixture is going to be an immunoassay, and that's not going to give you any sort of quantitative count in any sort of "field" fashion...

...but... just because there's actually a microscopic chance of pulling this off somehow, can you give specific drugs you're looking for? 'cause I can promise you, there's no chance in hell you're going to be able to run a blank check for "droogz" and get a one-size-fits-all solution. 'n rare p450 disorders are probably going to throw off any estimate which does come up, but shit like this is basically going to be desperately targeting one-specific-metabolite at a time...

Fair warning. Even that's a long shot, kid... but... fuck it, I'm bored, and attempting the unrealistic is something which is very good at wasting time.

reminds me of someone who wrote " fuck the shut up"

I support this proposal. Agoraphobic gurlz need luv too. ;)

i hate to spoil the OP's party, but not self respecting lab is going to adopt a testing regime without a published paper, or at least a paper that could be published. There is a very very long list of things you need to establish, not limited to:

known selectivity with reference to all likely interferences.
known limits of detection
confidence in results
lifespan of any reagents (including bench life, freeze thaw cycles and freezer life)
internal and external replications
available standards

making something like this is not difficult, i would take all the money you are spending on a lawyer and hire a chemist. you may have an awesome technique (please share it if you do) but no one is going to spend money on it without alot of convincing.

I've actually found a way to isolate a few hormones and 3 other substances in a manner in which they produce a concentration change in a separate compartment. If that seems really weird it is, but it works very well.

I've also found that a standard TDS meter will work with one tweak to my previous design.

The basic principle my device works on is similar to a pregnancy test, but it can distinguish between high and low levels of several hormones.

But JoePedo was right, I did a shit job explaining this.

EDIT: I was looking for ways to find several growth hormone and steroids mostly. I do include a VERY broad test for infections in the test array as well. I'm testing for 12 substances in all, but this is just a prototype so I can sell the patent.