Sciencemadness Discussion Board - Home Made Analysis gear

Home Made Analysis gear

aga - 11-4-2014 at 13:37

What kind of stuff do the Pros use to analyse chemicals, and Why cannot that equipment be made by home-brewers ?

I have heard of Gas Spectrometry.
I assume that means zapping a specimen material and analysing the spectra of the resultant gas.

What analysis kit is actually useful, and why cannot it be home-made ?

aga - 11-4-2014 at 13:41

I saw a TV programme where Australian Border Control had a machine that picked up on any trace of a controlled substance really fast, using just a swab of the suspect container, not touching the actual chemical material itself.

How does that work ?

I would very much like to make one, but for all materials, not just drugs.

E.G. : Is this Benzene ? Put sample in machine ..... Beeep! No. It's water.

gdflp - 11-4-2014 at 13:58

Quote: Originally posted by aga

I saw a TV programme

There is your answer. Most things used to analyze chemicals are way beyond the reach of an amateur, they require special machining to ensure they are precisely cut. Other equipment includes gas chomatographs, spectral analyzers, spectrophotometers, and NMR's(which are about .5 million per).

aga - 11-4-2014 at 14:14

Where there is a will, there is a way.

Machining : i was born in Sheffield. Machining to 0.001 mm is done routinely these days.
Etching to near-molecular tolerances is also a modern fact.

Gas chromatograph ? So measure the spectra of a gas. Do-able.

NMR i have no idea what that is.

OK.
That's some very good reasons why Not, so suggestions about 'How To' welcome.

[Edited on 11-4-2014 by aga]

aga - 11-4-2014 at 14:18

If the several brains in this forum could actually cause a home-brew Substance Identifier (hopefully quantifier as well) to be Possible, wouldn't every chemist want one ?

I would.

DrAldehyde - 11-4-2014 at 14:22

Quote: Originally posted by aga

If the several brains in this forum could actually cause a home-brew Substance Identifier (hopefully quantifier as well) to be Possible, wouldn't every chemist want one ?

I would.

I want one, I propose we call it a Tricorder.

aga - 11-4-2014 at 14:28

OK. TriCorder is the project name.

[Edited on 12-4-2014 by aga]

hyfalcon - 11-4-2014 at 14:54

These multiple posts are going to get you no where but detritus.

gdflp - 11-4-2014 at 15:08

NMR = Nuclear Magnetic Resonance. Have fun making one of those, they are very dangerous, magnetic fields and radiation and such.

aga - 11-4-2014 at 15:09

Sorry for the multiple posts.

Alcohol and the inherrent excitement about a project to make something good are to blame.

Mainly the alcohol.

[Edited on 12-4-2014 by aga]

aga - 11-4-2014 at 15:11

Thank you gdlfp for explaining, to some extent, NMR.

HgDinis25 - 11-4-2014 at 15:24

C'mon aga here is bringing an important topic that hasn't been done yet. So stop saying you can't and let's start brainstorming to get somewhere.
hyfalcon, with all due respect, your post deserves much more to go to detritus. You didn't post neither to help nor to give new ideas, you simply post to warn about a little mistake (doublepost). Everyone here could see that, so if you are the "obvious pointer" guy, go point the obvious somewhere else. Otherwise, quit that attitude and say something interesting.

Now, on topic. I've made some CD spectometers just for fun. You can easily see some spectrum lines. I wonder if, by placing a sample in the spectometer, we could analyse it's spectrum? It seems doable to me. For those who don't know, here is one:
https://www.cs.cmu.edu/~zhuxj/astro/html/spectrometer.html
I prefer this aproach on making it:
http://www.youtube.com/watch?v=D_gT2dl3B5Q
I'm out of time to test it out, but as it is really simple to make someone here could test it?

Density tests using a Pycnometer and a good analytical scale are also pretty acurate determining a sample density. One could then compare it to a density table and look for your compound. Determining the melting or boiling point of a sample also helps do determine the compound and purity.
ChemicalMyst speaks about this in a video of his (right in the end of it):
http://www.youtube.com/watch?v=k7ZDSFEsrGg

Any more ideas to add on the table?

Chemosynthesis - 11-4-2014 at 18:14

Most forensic drug testing uses mass spec heavily. I'm most familiar with MALDI TOF. You need different systems for liquid, solid, or gas phase analytes. Flame ionization would probably be simpler than laser for LCMS (maybe), definitely GCMS.
http://en.wikipedia.org/wiki/Mass_spectrometry
http://en.wikipedia.org/wiki/Matrix-assisted_laser_desorptio...

I can't imagine making most analytical instruments because here you need precise timers, magnetic focusing and acceleration, an inert gas flow/vacuum, sensitive ion sensors, possibly a quadrupole. I can't even begin to give it justice. The combination of everything makes it harder than just spraying sample through a purged nichrome wire and hitting a screen.

My favorite analytical technique, and one also used in the drug industry frequently is NMR. NMR uses extremely strong magnetic fields under liquid helium, with an insulating layer of liquid nitrogen. This is expensive, hard to maintain, can lead to very difficult to diagnose issues, and is dangerous if you get a "purge." It requires dedicated space away from all magnetic object, computers, credit cards, and the software to run them tends to be licensed and is designed to run on Solaris/sunOS UNIX on every system I've used (Varian). It is also extremely sensitive, so you need very pure sample, a standard (generally TMS), deuterated solvent, and very fine calibration to compare spin-anisotropies to get Knight and chemical shifts through perturbing the nuclear spin states. It's pretty complicated, I've forgotten a lot about it (except how to read some of the spectra), and I was never considered an expert in the technique. People get paid fulltime to do these kinds of things, including mass spec... and I never was. It's so sensitive (ppb/ppt), there were plenty of times I had to rotovap material repeatedly and high vac overnight to get a clean spectra without extra solvent peaks. Maybe not necessary for home science, or even all publications, but it can be annoying.

HPLC is also complicated. You have special columns, pumps, etc. I think you'd be best off faking this with flash chromatography after TLC, but I could theoretically see someone plumbing this together. Won't be a Waters or Agilent UPLC, but I could see it, maybe.

Now you move into things like IR spec, Raman, and UV-Vis. I could actually see you making these, though I have seen FT-IR for hundreds used, and Raman for a grand. At the very least, you can easily get these in your choice, used for a couple grand any day of the week, so I don't see why you would make one unless for fun.
http://www.sciencemadness.org/talk/viewthread.php?tid=23422
http://www.youtube.com/watch?v=-lVMYa25jtU
doi:10.1021/ed800081t

All in all, though machining is amazing, don't underestimate how electronic a lot of this will be, including calibration and maintenance. I do want to emphasize I am impressed by your machining background. Don't laugh, but I bought a CNC Taig micro mill (the larger one... can PM model numbers), thinking I would learn how to use it, and never found the time or anyone willing to teach me, so I am fascinated. I would love a Sheffield and some experience!

Edit- are you more interested in the fun and accomplishment of building your own, or just the end result? A lot can be done with melting point analysis, rf values in various solvents, and even a density estimation, as well as various chemical characterization techniques for functional group presence.

[Edited on 12-4-2014 by Chemosynthesis]

Dr.Bob - 11-4-2014 at 18:48

Some instruments require very complex electronics, complex gases, cold temps, or expensive parts. But if you look through the other parts of this forum, you will see discussions on home made, portable, or smaller NMRs, home made GC and HPLCs, and polarimeters. There are smaller versions of many instruments which could be adapted for home or hobby use. The reality is that if you could build most of these items for cheap and they worked well, someone would be working on it already. But some instruments like CG and MS have been miniaturized recently and are much smaller and less expensive than they were just a few years ago, I would not be surprised if that trend does not continue.

That may be true of analytical tools, but preparative equipment, larger scale instruments used for purification of larger amounts of material, where the purified material is the product, not the analytical information, are still expensive in general, like prep HPLC, flash chromatography instruments (although they have dropped in prices due to being more common now.)

High quality NMR is still expensive, mostly due to the cost of the superconducting magnets used for most higher quality machines, but if anyone can make higher temp superconductors work for NMR magnets, then that could make a huge impact on NMR, as the magnets often cost $100,000's alone, and the electronics (like radio freq. probes and ppm signal amplifiers) are very complex and not trivial to build cheaply, but that also may change also.

Bert - 11-4-2014 at 20:44

If we start to see some hard references and , this is going to stay. I really hope that happens, I've seen "amateurs" and "makers" pull off things pros are hard put to do, and on a shoe string.

If it continues as blue sky, but has educational links to information on instrument design & theory, it's beginnings... Already enough info here to take it past detritus.

Whoo hoo! 30 + years ago, I got to help build an NMR.

[Edited on 12-4-2014 by Bert]

aga - 12-4-2014 at 00:17

Wonderful !
I knew it was worth waiting for some Enthusiastic Thinkers to turn up.
Thanks for the references - i will follow up each of them.

I suppose to be practical, the machine needs to characterise as much of the substance is is feasible.

Complex electronics and computer software are not an issue for me, neither is a modest degree of cost.

It sounds like NMR will not be an option, due mostly to the Hazards and immense cost.

Currently i'm thinking along the lines of :-
Freezing point
Melting point
Boiling point
Spectral analysis, including visible, IR and UV.
Without knowing what frequencies might be required (will look it up), RF resonance may be useful.
Then there's Excited/Activated state responses, such as spectra after zapping with 30kV, spectra after exposure to the collapse of a large magnetic field etc etc.
Piezo electric effect ? - maybe some substances squeak (in some sense) when exposed to high pressures, or a vacuum.

These all spring to mind because they should be measurable with electronic sensors.
Things like solubility in solvent x, y and z would be much harder to do electronically.

[Edited on 12-4-2014 by aga]

aga - 12-4-2014 at 00:50

Quote: Originally posted by Chemosynthesis

I bought a CNC Taig micro mill (the larger one... can PM model numbers), thinking I would learn how to use it, and never found the time or anyone willing to teach me, so I am fascinated. I would love a Sheffield and some experience!

I found this to be excellent, and should help you to get your Milling machine cutting metal :

http://littlemachineshop.com/Info/getting_started.php

A first task would be to mill a straight slot in a piece of aluminium or brass (they're soft, so tolerate mistakes better and are less likely to snap your tooling).

hyfalcon - 12-4-2014 at 01:58

My only problem was the rapidity of his posts. If he has another thought that quick then he could just edit the previous post instead of making post after post. Look at the time stamp on the first posts.

And I'm sorry, the OP lost me with "My ass".

[Edited on 12-4-2014 by hyfalcon]

Bert - 12-4-2014 at 05:23

Quote: Originally posted by hyfalcon

My only problem was the rapidity of his posts. If he has another thought that quick then he could just edit the previous post instead of making post after post. Look at the time stamp on the first posts.

And I'm sorry, the OP lost me with "My ass".

[Edited on 12-4-2014 by hyfalcon]

Agree. I got to this late last night from a SPAM report, caused by that repeated posting. Also, "My ass" might JUST be suitable for whimsy, if you smile while you say that, Pardner! Definitely suitable for Hot Electron Death otherwise.

aga:

Please go back and edit/remove your earlier posts, for civility and content? Make it something worth keeping around- you've got 24 hours.

The first time I saw one of these, I have to admit I DID think "TRICORDER! WHOOT!!!"

neptunium - 12-4-2014 at 05:26

how about neutron activaction?
depanding on the cross section of your sample you can identify metals (mostly)
How about XRF ?
withe the same equipment used for neutron activation plus a xray tube your in business!!!

http://www.sciencemadness.org/talk/viewthread.php?tid=27963

hyfalcon - 12-4-2014 at 10:39

XRF, That's what I was trying to remember and couldn't.

http://goldrefiningforum.com/~goldrefi/phpBB3/viewtopic.php?...

[Edited on 12-4-2014 by hyfalcon]

packetforger - 12-4-2014 at 10:54

I always did wonder why no one had thought to try make a homebrew IR spec. or UV-VIS setup.

AFAIK the main problems would be obtaining IR/UV sources and sensors which could emit all the different wavelengths and detect them with high degrees of precision and accuracy. After that, it would simply be a case of analysing the output data.

I actually know how most of these machines work inside (and if anyone wants proof of me being a capable analytical chemist, I can now provide it!), the main issue for the homebrewer is obtaining the detectors and emittors of various wavelengths/signals IMO. That and the tight tolerances, etc.

How a UV-VIS works, for example, is a very basic principle. Shine a certain wavelength through a sample, see if any of the light is absorbed. Record how much is passed through compared to how much was emitted on one side. Try this with a bunch of wavelengths of UV light. Different materials absorb at different wavelengths.

A = E.C.L where A is absorptivity, E is the molar absorptivity coefficient, C is the concentration of the material (moles per decimeter cubed) and L is path length of the cell, normally 1 centimeter. That is the basis of that.

An NMR machine would require extremely precise sensors, radio-wave sources, and strong electromagnets, among other kit. I doubt someone will be making one of those anytime soon.

If you want your samples analysed, your best bet IMO is to make friends with a chemistry student at a nearby Uni with a good lab. Most of us have no problem running a few samples out every now and then, and our lecturers often tend to be somewhat supportive of us getting more practice with the techniques.

Perhaps some genius will find a cheap source of IR and UV sensors/emittors capable of the frequency ranges needed with decent accuracy/precision, knock up a bit of software (probably Arduino based) and put up a kickstarter someday...

Edit: as for multiple-stuff-machines, I did once sketch out an "extension" for a HPLC machine that would take the fractions coming out (using a UV detector), do both IR/UV-VIS analysis on them to determine what wavelengths they absorb at (IR is useful for obtaining functional groups and suchlike, UV I have only ever used for concentration data, but it may come in handy), then pass it along to a little collection device that would neatly collect each fraction. A complete pipe-dream device, one I lack the expertise to build, but was fun to come up with. You could then run NMR and MS on the samples to gain more information.

[Edited on 12-4-2014 by packetforger]

Etaoin Shrdlu - 12-4-2014 at 11:38

I've never been much for mechanical work, but I have a lot of experience working with spectral data and quite a fair bit of programming knowledge. I recently cobbled together an excel spreadsheet with programming tagged on* to analyze the output of our visible-light spectrophotometer, express how a material appears under different light sources and through different receptors, etcetera, with a high degree of customizability for non-real-world scenarios, and I'm currently building an add-on to see if I can get it to pick up on specific patterns and output likely pigments/dyes contained in a sample, based on their spectral data. I can simplify readings to a "pure" color curve based on the spectrum of the light source, and (theoretically) known absorbance from the container (edit: or based on a reference sample if that's available). I think a lot of this could be adapted to UV/VIS/IR spectrometry, and all of it can be transferred to something standalone if necessary, though I'd have to learn how to get a program to talk to the machine, or have some way of retrieving data through an intermediary. So if anyone does try their hand at this, I'd be interested in offering assistance on the software side of things, however fumbling it may be.

*(Oh I know this would be better served by a separate database and program, but Excel allows me to have both together without fussing with compiling code or compatibility issues. Shhhhh.)

[Edited on 4-12-2014 by Etaoin Shrdlu]

aga - 12-4-2014 at 12:20

Quote: Originally posted by packetforger

I actually know how most of these machines work inside (and if anyone wants proof of me being a capable analytical chemist, I can now provide it!),

Invaluable knowledge !
Do you also have access to 'proper' tst gear so that any results could be checked ?

(assuming a TriCorder gets built)

The incredible work done by radagast last year looks like a very promising place to start :-

Quote: Originally posted by radagast

Amateur chemists need a way to identify and characterize compounds

[Edited on 12-4-2014 by aga]

aga - 12-4-2014 at 12:24

My apologies hyfalcon.

I did not intend to offend anyone.

Now i have a yardstick as how easily that can happen, i will make an effort to moderate my posting behaviour.

--------------

To avoid annoying anyone further, i am editing this post rather than posting a new one.

I do not feel that i will be able to avoid annoying people, as i do drink, think and post.

As new ideas crop up (in my mind) i find that sending them out as-is for discussion works well, as people react, flame, argue, and contribute.

So, best if i do not post anything further.

Thank you all for your very informative and detailed answers to questions i have asked, and please continue supporting beginners in the future.

Kind regards,

Adrian.

[Edited on 12-4-2014 by aga]

Chemosynthesis - 12-4-2014 at 13:42

Quote: Originally posted by aga

Currently i'm thinking along the lines of :-
Freezing point
Melting point

I could have sworn someone at least started their own meltemp apparatus here, but can't find the post.

http://www.chemicalforums.com/index.php?topic=50619.0
https://www.sciencemadness.org/whisper/viewthread.php?tid=22...
https://www.sciencemadness.org/whisper/viewthread.php?tid=16...

Those would be useful. A meltemp can't be all that hard to make. If you had a small camera with the temperature and sample in frame, you could leave it to slowly run and get fine measures.

Wouldn't be as cheap as a Thiele tube, but much more time convenient.

Quote: Originally posted by aga

Wow, thank you! Maybe I can get it out of storage this summer. I am very appreciative!

Quote: Originally posted by aga

I do not feel that i will be able to avoid annoying people, as i do drink, think and post.

As new ideas crop up (in my mind) i find that sending them out as-is for discussion works well, as people react, flame, argue, and contribute.

So, best if i do not post anything further.

I hope you reconsider. Between your machining and electronics skills, packetforger's analytical chemistry education and background, and Etaoin Shrdlu's programming abilities, I was very hopeful for some extremely insightful experimentation.

packetforger - 13-4-2014 at 05:27

Once my exams are done or I get some time to transcribe and redraw diagrams, I will see if I can sketch out some decent representations of how some of these machines work at a fairly basic/conceptual level.

If a variable-wavelength UV source could be found (i.e. one where we can accurately control the wavelength emitted) and a detector which can identify both wavelength and intensity, a fairly simple UV absorption spectrophotometer could be constructed, which is *very* useful with a set of known standards for determining the concentration of an unknown material (although, in many cases titrimetric determination is more accurate!). Identification can be also aided by finding the peak absorption ranges on the spectrum of a compound.

Similar in principle is the IR Spec. Now, I have seen two varieties of these in the lab. The first, the one I like using the most, you simply place the sample on the "crystal" which has been cleaned by wiping with acetone, press down the "cover/crusher thing" (I don't know the term for it, we called it the "arm"), and clicked on "scan" on the computer after running a background.

I suspect that a homemade one would have two sample areas, one for a "blank" (container containing only the solvent), and the other containing a solution of your target material. The blank's peaks would be factored out, and the "true" IR absorption spectra of the compound recorded on a graph for further identification. IR is great for spotting functional groups (determining if its an alcohol or a ketone or an aldehyde, etc), and generates a "fingerprint" that can be used to identify a previously known compound based on libraries of such.

Edit: Obviously, the only massive obstacle here is the variable wavelength sources and detectors. Fix that problem and we have a winner.

[Edited on 13-4-2014 by packetforger]

Mildronate - 13-4-2014 at 09:01

you can build mercury polarograph. But actualy you need analitycal balance for real analitys chemistry, its also posible to build autotitrators and other stuff.

packetforger - 13-4-2014 at 10:31

Mildronate is correct, one of the more important tools at your disposal will be an *accurate* analytical balance that can weigh down as far as 4 decimal places with an incredibly high degree of accuracy.

Not to mention, ensuring you have it placed on a decent surface is equally important as a mass-dampener. The ones in one of the labs in college are not exactly the best positioned, and other people working at said area can throw your readings off by a significant amount. For example, someone leaning on the surface your balance is working on can alter the perceived mass the balance is detecting, etc.

Ideally you would place your balance on top of a big, solid lump of something heavy. A solid wooden table with some weights on it or a big slab of stone or something.

ScienceHideout - 13-4-2014 at 15:44

In my experience, the most important analytical device in the lab is the analytical balance. You must buy these... For the home analytical chemist, milligram capability is important; tenth milligram is better.

The most useful spec for the home chemist is probably the UV/Vis. I own one of these that works fine, except for the computer connectivity which is being worked on so I can do scans. Honestly, you can probably build one of these, or, at the very least, use a whole bunch of LEDs instead of a monochromator.

The most useful spec for organic is the IR. It is easy to use, and will analyze almost anything with a couple presses of a button. I want one of these, but they are super expensive.

Gas chromatographs are used to separate mixtures of liquids/gases. I am actually building one of these and it will be complete this summer, when I will post a set of instructions on the forum. Many GCs are connected to a mass spectrometer... sorta useless to most home chemists, but could come in handy. They are advanced...

HPLC is similar to GC. You can probably build one by connecting a column to a UV/Vis spec and a flow cell... Never tried it.

Flame ionization seems to me, to be one of the most useful specs... and pretty cheap on ebay. Unfortunately, They are huge and I doubt I can fit one in my lab.

NMRs are cool, they measure how the molecule reacts to radio waves, and that will show all the hydrogens in the molecule. That means, though, that your sample must be dissolved in a deuterated (expensive) solvent, or else you will be in for a surprise! Also, these are out of the question for home chemists. Not only will they demagnetize all the credit cards in a large radius, but they cost upwards of $100K, often around 250-500K.

I think the most amazing instrument, and I have been intrigued by it ever since I have used it, is the EDAX spectrometer... THese are amazing! The only problem is it has to be connected to an expensive microscope

kavu - 14-4-2014 at 01:40

You can get these new low field tabletop NMRs with permanent magnets and a flow cell. With some savings the price range is actually within the reach of an amateur. Check out for example http://picospin.com/

[Edited on 14-4-2014 by kavu]

Lambda-Eyde - 14-4-2014 at 01:54

Benzylchloride1 also has an NMR, an old permanent magnet from Hitachi iirc.

Quote: Originally posted by ScienceHideout

IMany GCs are connected to a mass spectrometer... sorta useless to most home chemists, but could come in handy. They are advanced...

What? In addition to IR, I can't think of anything more useful than GC/MS!

Mildronate - 14-4-2014 at 05:00

But what do you want to analyze? There is full of old equipment around the world. Why to build some? If i ever will build GC, I will use real GC parts! but i doesnt need it, because I working with GC and HPLC.

neptunium - 14-4-2014 at 06:07

i have worked before on trying to build a mass spectrometer to me thats the holy grail of analytical instrument.
However ity requires deep vacuum and very god ion source and magnetic field it might be easyer than it sound with all the new electronic replacing the old ,parts could be purchased cheap or home made without costing thousands...

ScienceHideout - 14-4-2014 at 11:13

Quote: Originally posted by Lambda-Eyde

What? In addition to IR, I can't think of anything more useful than GC/MS!

To be completely honest, when we are talking about liquids, I can't think of one thing a GC-MS can do that an IR can't... Mass spectrometry is so much more difficult to use and interpret, they are more expensive, and they take up so many more resources. If you give me an unknown mixture of 2 liquids, I would pop it in the GC, condense the condensate, then IR it...

aga - 14-4-2014 at 13:02

Quote: Originally posted by Chemosynthesis

Ok. I was just pissed off with the initial responses.

It is very hard to be enthusiastic when the Enemy Fire starts before you even leave camp.

However, the responses Since are brilliant !

It's also My thread, seeing as i started it, and i should defend it, given that the Idea is to make an Affordable Analysis Machine, and that You People have the knowledge to make it work.

I've ordered some stuff off ebay to build a 'monochromator' which in principle is rather simple, and found two 523nm lasers in a drawer.

The Volatile Chemist - 14-4-2014 at 13:03

A little while ago, I devised a system to tell the frequency of emission of light, specifically IR or visible, with two or more DIFFERENT light sensors. For example, a CdS photo-resistor, and some other type that overlaps the range of CdS are used to measure light output (Strength, as with all photoresistors and (Extremely!) cheap sensors). These two sensors have a range of how well they pick up certain frequensies, which is commonly published. With two bell curves from this data, we can find which frequency is being emitted by finding what the strength readings are for each sensor in a given situation.

And have you checked out EPR?
http://en.wikipedia.org/wiki/Electron_Magnetic_Resonance
For example, do I have CuCl₂? It emmits an EPR, but I don't know how hard such a machine would be to build.

[Edited on 4-14-2014 by The Volatile Chemist]

aga - 14-4-2014 at 13:08

Another interesting factoid is that the 10Ghz range of frequencies are useful in excitation and analysis.

This immediately sprang to mind :
http://www.qsl.net/pa3gco/zelfbouw/blauwkap/bluecap.html

Basically the LNB part of a cheap satellite setup contains very high frequency amplifiers, and an oscillator at 10Ghz, plus two (Horizontal & Vertical) precision antennae, plus two amplifiers as well.

It all comes in a neat, and frequency tuned, aluminium case as well.

Cost = 16 euros here for a modern Sony version, which i took apart and found that it can also be modified.

[Edited on 14-4-2014 by aga]

aga - 14-4-2014 at 13:29

Quote: Originally posted by kavu

You can get these new low field tabletop NMRs with permanent magnets and a flow cell.

Amazing. I was doing some research into the Magnetic Field Strengths required for Mass Spectrometry and 0.35 Tesla came up.

Seems a permanent magnet can possess such a field, and if not, Halbach arrays, or even spheres can be employed.

The AA batteries in this TriCorder are going to get hammered.

aga - 14-4-2014 at 13:32

Quote: Originally posted by neptunium

i have worked before on trying to build a mass spectrometer to me thats the holy grail of analytical instrument.

Please elaborate. It could be that one of the stumbling blocks of a previous attempt of yours can today be simply solved.

Certainly Electronics advances faster than money leaves my pocket.

Etaoin Shrdlu - 14-4-2014 at 13:36

Quote: Originally posted by kavu

You can get these new low field tabletop NMRs with permanent magnets and a flow cell. With some savings the price range is actually within the reach of an amateur. Check out for example http://picospin.com/

[Edited on 14-4-2014 by kavu]

That is a beautiful thing.

aga - 14-4-2014 at 13:42

Quote: Originally posted by The Volatile Chemist

A little while ago, I devised a system to tell the frequency of emission of light, specifically IR or visible, with two or more DIFFERENT light sensors. For example, a CdS photo-resistor,

I have a feeling that Cadmium Sulphide cells will not be sensitive enough.

Personally i am thinking more along the lines of USB webcams (or similar) which are generally CCD, cheap, and sensitive to both visible and IR spectra if you remove the IR filter.

[Edited on 14-4-2014 by aga]

leu - 14-4-2014 at 13:59

You can use various derivatives, paper or thin layer chromatography and a pc/flat bed scanner or possibly an Android cell phone to do do quantitative analysis:

http://www.sciencemadness.org/talk/viewthread.php?tid=5991

There's a copy of Vogel's in the forum library which will explain the techniques in great detail :cool:

The Volatile Chemist - 15-4-2014 at 06:50

Webcams are a great Idea! Not perfect, but possibly any digital camera would work! I have an old one (Battery case broke, but when you plug it in, it's fine), so I'll experiment with removing filters. Does anyone know where I could find find a tutorial of how to remove the filter. I know what aga said works, because I saw an instructables article once about making a thermal camera out of a digital one. Obviously not perfect, but it would work.

I have a fire sensing diode, an IR sensor painted black. I got it a while ago, but I wonder if such IR sensors would work, or if we could paint such a camera black to get the same effect for a camera.

If we get this working w/ a camera, TSView (Software) would be perfect. Import a BMP you took from the camera, use the pseudo color function to bring out differences, etc. or just make a program to get the average pixel value to find out an average frequency. It'd be a start at least.

confused - 15-4-2014 at 07:11

any webcams that can do uv as well as visible light?
If so, a uv-vis spectrometer could work

Volatile Chemist:
http://www.instructables.com/id/Making-a-Night-Vision-Webcam...

The Volatile Chemist - 15-4-2014 at 08:28

Quote: Originally posted by confused

any webcams that can do uv as well as visible light?
If so, a uv-vis spectrometer could work

Volatile Chemist:
http://www.instructables.com/id/Making-a-Night-Vision-Webcam...

A few people have stated they can do IR and obviously they can do visual. It seems reasonable they could do a bit of UV too, but not much, and it's likely there'd be no way to filter visual spectrum out from it.
Good Idea though! When I get home from school, I'll take apart my old digital, and see what I can do in the means of filter removing

aga - 15-4-2014 at 23:51

Right now i'm waiting for delivery of the components to make a Raymen analyser based on Radagast's work, with an idea i had for the monochromater, and a camera module as the detector.

The mel-temp part could use the same camera, and will be a block of aluminium with a hole in it for the capilliary tube (some basic machining required) and a soldering iron 50W heating element + temperature control circuit.
The block of ali should help make the temperature changes smoother.

I like the thin-film chromatography idea as well, seeing as it could also use the camera, measuring the speed & distance at which bands separate.

Any clues as to the Length of the thin film section ?

Would two slides clamped together be long enough ?

hyfalcon - 16-4-2014 at 03:24

Quote: Originally posted by The Volatile Chemist

Quote: Originally posted by confused

any webcams that can do uv as well as visible light?
If so, a uv-vis spectrometer could work

Volatile Chemist:
http://www.instructables.com/id/Making-a-Night-Vision-Webcam...

Check into Wood's glass. It filters out visible spectrum bandwidth and passes UV.

Mildronate - 16-4-2014 at 06:02

How you guys ever worked with real MS or NMR? There always is many problems, homemade NMR or MS its not reality. And there is always classic methods for molar mass determination like cryoscopy or ebuloscopy, refraction and other more useful in home, first try basic and then think about MS. How often you need MS? I working on GC with MS, but i use more often FID for analysis. I also had HPLC at work with MS-MS, but if i can i chose FDD, use MS only for amino acids quantitative determination.

dontasker - 16-4-2014 at 17:58

Slightly off topic but still relevant:
http://benkrasnow.blogspot.com/2011/03/diy-scanning-electron...

He's also done freeze-drying, super critical CO2 extraction, a CT scanner, and much more.

Check out the YouTube channel he has. Amazing to watch and he brings it down to a level even I can understand.
https://www.youtube.com/user/bkraz333

neptunium - 16-4-2014 at 18:51

Quote: Originally posted by Mildronate

How you guys ever worked with real MS or NMR?

NMR ? no but MS yes

have you checked that one out ?

http://www.sciencemadness.org/talk/viewthread.php?tid=23422&...

DistractionGrating - 17-4-2014 at 11:14

Quote: Originally posted by aga

Personally i am thinking more along the lines of USB webcams (or similar) which are generally CCD, cheap, and sensitive to both visible and IR spectra if you remove the IR filter.
[Edited on 14-4-2014 by aga]

I believe that CCD webcams have limited sensitivity into the NIR range, up to only about 900 - 1100 nm. Most NIR specs cover a range beyond that, more like 1100 nm up to around 1700-1900 nm. Usually InGaAs, Ge, or PbS detectors are required for that wavelength range, as far as I can tell.

I'm new around here, but have a strong desire to build a NIR spec. Please see my newly created thread about it here: DIY NIR Spectrometer?

aga - 17-4-2014 at 12:45

Specs on the sensors used in cameras are hard to find, as they tend to just say 'colour' and 'lo-lux' etc.
I have a couple, and will try to find some markings on them to identify the chip, and then try to get some specs.

There was one time i ordered two cameras (for FPV), and the supplier asked if i wanted the 'colour correction infra red filter' removing, which suggests that IR sensitivity exists in these units, and a filter is needed to correct the colour responses.

Can i assume that NIR gets refracted by a prism ?

Update on Sources of Parts :-

I took a 3.5" hard drive apart today, and punched down thru a hole to see what happened.
After full disassembly, it seems i broke one of the two platters.
It is Glass (or some ceramic that shatters like glass).
It is also coated with a metal substance, is extremely Flat, and most likely mechanically stable over vast temperature and acceleration ranges.

It's a 'perfect' mirror as far as i can tell.
Stick it next to a surface and there is no perceptible 'ghosting' or colour difference in the reflected image when compared to the image seen normally.

The chassis is anodised aluminium, but is rather small for my idea for the mel-temp.
A larger, older (5") hard drive chassis will probably serve well.

2 for 1 week in the electro-garbage if this idea works !

(also a 'no reserve' auction for 20 dental size #5 front-coated mirrors)

Baffled - 20-4-2014 at 03:04

A good old Thiele tube goes a long way, if you have a accurate scale you can even calculate the molar mass (with a certain margin of error depending on your scale) using freezing point depression:
http://www2.vernier.com/sample_labs/CWV-15-COMP-freezing_pt_...

I have one of these: http://potassium.1338.at/upload/pictures/schmelzpunktbestimm...
The silicon oil (some high boiling PDMS) that goes with it is terribly expensive, but it works!

Mildronate - 20-4-2014 at 04:11

Yes its crioscopy there is also ebuloscopy (boiling point depresion). But there is only one problem you must have very clean chemicals.

aga - 28-4-2014 at 12:17

Still waiting for parts to arrive. Spanish Postage is not so fast.

I was wondering if this thread should be closed, and any useful efforts moved over to Radegast's thread, seeing as Rad actually has something built and working, whereas this one is still just talk.

What i will attempt to build is an accurate spectrometer using cheap parts (when they arrive) which will be computer controlled and self-calibrating if possible.

The detector will be a ccd camera of some sort, and as i have seen mention of time-synchronisation being useful somewhere, in relation to irridescence/flourescence, my current thinking is to synchronise the laser pulse with the line-synch pulse of the camera, which is once every 64us. Clearly the laser pulse duration and time offset will need to be varied and experimented with.

The notion of a polarimeter is interesting, so i will include the possibility of analysing the reflected, but polarised laser light reflecting off the sample, passing through a rotationally-adjustable second polarising film.

Looking at the constructional possibilities, i would hope that this will end up as a hand-held unit which can be pointed at a substance, and read out the composition, although i appreciate that will be a Tall Order.

The Melting Point aparatus seems trivial. USB microscopes are available for a few $, so all that is really required is a heating element, a block of aluminium with some holes in it, a digital thermometer and a temperature/heating controller, so i suggest that it be spun off as a separate project, unless anyone can suggest a valid reason that MP would be a clincher in determining the composition of a substance.

My only objection to MP is the sheer energy required to raise the temperature, and so not a hand-held device.

Does anyone have any suggestions regarding the quantitisation of the detected substances : i.e. work out how much of each there is ?

[Edited on 28-4-2014 by aga]

Mildronate - 28-4-2014 at 15:28

You need standart solutions for quantitative analysis, analitycal balances and A class volumetric flasks 10, 15, 25, 100, 250 mL more than one of each.

Blue Matter - 30-4-2014 at 12:18

Semi off topic but you might find this interesting.

https://www.kickstarter.com/projects/903107259/scio-your-six...

aga - 30-4-2014 at 13:05

Bang on topic !

That's amazing.

Game On i guess. A Prism & some lenses arrived today, along with the heating elements for the MP thing.

Etaoin Shrdlu - 30-4-2014 at 14:34

Quote: Originally posted by Blue Matter

Semi off topic but you might find this interesting.

https://www.kickstarter.com/projects/903107259/scio-your-six...

I really don't like that they won't show the spectra. Why? That should be something they are eager to allow people access to (and also the simplest part of the data to grant access to).

Otherwise, that device is quite cool. So small and accessible, too. I'm excited to see how this plays out.

Blue Matter - 30-4-2014 at 16:04

"We will consider providing access to the raw spectra to specific users according to their needs."

I suppose if you shell out the cash you can get a model that shows the raw spectra?

aga - 3-5-2014 at 11:27

SCIO Pro i guess @ $15,000

Mildronate - 3-5-2014 at 12:58

I actualy interesting about polarograph building, its very interesting method and more possible at profesional level @ home than diY FTIR or NMR

Blue Matter - 3-5-2014 at 14:57

out of curiosity how many people are actually planning on building one in this thread?

Mildronate - 4-5-2014 at 00:22

I am planing polaragraph

Actually i have polarograph, but i doesn't have droping mercury electrodes for it, mercury is not a problem, problem is electrodes design.

The Volatile Chemist - 6-5-2014 at 06:26

I'm interested! At least a spectrophotometer of sorts.

aga - 6-5-2014 at 13:24

measuring the spectral response is the key.

all the rest is applying a stimulus and factoring it out of the reading.

So, spectrometers all the way..

Brewing Analysis

The_Taxidermist - 9-1-2017 at 03:13

Interesting topic, and responses. Being new, and a subject dear to my heart, I note that there is an element of discussion missing, which I can fill in. What type of analyses for brewers/homebrewers are there (should there be) in use?

Malt Analysis, provided per request as a sheet from the malt supplier.
Water Analysis, as provided, usually, by the Supply Company.
Sensory Analysis, as evaluated by the brewer and friends.
Colour Analysis, also as provided by brewer and friends.

For the (uber-keen and closet "mad scientist" in us all) hobbybrewer, this site outlines various instrumental techniques for analysing beer; Advanced Analytical Methods...for Beer

Beer Analysis (Molecular method of plant analysis) The Index to the book gives a good listing of various analyses;

and for Beer color determination, this from Lovibond Tintometer

Beer Colour is mostly judged by eye using beer color swatches and takes some practice and training to learn. Lovibond made a series of glass filter disk standards (€137/disk 4 to the complete set) for their Lovibond 1000 Comparator system (now more or less superseded by newer tech, but still supported) and was aimed at laboratory use. I have one of these but not the specific disks.

From my observations, analysis for bitterness and alcohol concentration would be desirable goals. NIR for fermentation analysis would be cool. We already have specific gravity or refraction for sugar conc. Do we want to analyse how viable our ageing hops are? You bet! Do we want to check the quality of our 1-2 year old (rodent/pest free) grain storage? Of course. But as a first step, for me, final beer colour is potentially accessible.

I myself have been slooooooly working on a 413nm transmission colorimeter that fits into the size of an iPhone 5 size-like packaging box, I have most of the components an idea for the circuit but am missing one key element - basically I can't decide on the right Log Amplifier and am suffering sticker shock around the one I thought would be best.

Also in short time, I'll be pulling up sticks and moving to Shanghai so al current project have been put on hold.

Good luck with this concept? I'd love to see more ideas about what we can hobbybrew-erise into stand-alone labtop instruments, that don't cost a fortune, are easily diy-accessible, and produce satisfactory results – think 'food grade' reagent, as opposed to 'Analytical Reagent' grade. AR is great, don't get me wrong, but as a hobbybrewer, FG would be exemplary, if I could only actually get it.

Last point, some of the Kickstarter ideas getting around, rely far too much on proprietary clouds to drive them. Whilst being networked is a nice idea, I don't think I need some cloud system to drive an instrument, analyse the output, or store it for my personal data safety.

That's a bit like the doctor emailing me the results of my latest stool analysis when I didn't even know they'd already been storing and logging the samples.

I think I can manage quite well without the implied beneficial extras. Simple instruments with a readable display, and a lab book. What ever happened to the K.I.S.S Principle?

Cheers.

[Edited on 9-1-2017 by The_Taxidermist]

[Edited on 9-1-2017 by The_Taxidermist]

bobm4360 - 9-1-2017 at 11:44

Cason's"Laboratory Text in Organic Chemistry", 2nd edition; and Amateur Science column in "Scientific American", September, 1967, have detailed info on constructing gas chromatographs.
Regards,
Bob

tsathoggua1 - 22-1-2017 at 11:06

For comparative analysis, wouldn't a simpler way to go about mass spec, use of the calutron principle? only, separating ionized compounds rather than isotopic fractionation.

Might be interesting to try and build one even if it would be a shit for quantitative analysis (here, I am thinking comparisons via building up a sample reference library, tedious and not exactly cheap, but say one could (just for example, using familiar examples) scan benzene, halobenzenes, 2,5-dimethoxybenzene, 4-substituted 2,5-dimethoxybenzene, then the benzaldehydes, equivalent phenethylamines, amphetamines and cathinones (the last could be especially useful perhaps, and then scan similar ketones without the amine functionality, such as propiophenone and substituted propiophenones, or for that matter, halomethanes, haloethanes, ethers, to build up a library.)

Of course such an approach depends on having a reference library.
Although such for specific areas of interest could be built up with relative ease, and especially for solvents.

Chemetix - 22-1-2017 at 13:52

I have a soft spot for LC, I think that a back yard version could give some repeatable reliable performance. It's not HPLC and it's not going to give you the reliability of an alliance, but a low pressure pump some glass tube and a photodetector made from a pair of LEDs could make a fairly simple system. Some have some pretty sort wavelengths these days. Ok it's not 254nm but it would show some absorbance from the baseline. With a clean sample you might get the 3:1 signal to noise you need.

As for backyard reverse phase media...not calling this a success but I poked around with wax coated starch and packed a column with it. Caffeine gave a reliable curve from methanol water, the detector was a hacked Waters 486 with a germicidal lamp as a UV source. But there are so many detection systems that could be exploited.