Sciencemadness Discussion Board - Quick determination of ammonia or ammonium?

Quick determination of ammonia or ammonium?

blogfast25 - 16-9-2008 at 10:04

Does anyone know of a relatively easy way to quantitatively determine ammonia or ammonium? I'd like to apply it to my anhydrous manganese dichloride, where residual ammonium chloride may be causing me some problems...

DJF90 - 16-9-2008 at 10:14

Test the solution with sodium hydroxide solution. Insoluble precipitate may form (depending on the cations present), but this may redissolve on adding more hydroxide. Warming the solution will yield ammonia gas if the ammonium ion is present. Test with moist red litmus or HCl fumes.

EDIT - I didn't see the QUANTITATIVELY

[Edited on 16-9-2008 by DJF90]

[Edited on 16-9-2008 by DJF90]

not_important - 16-9-2008 at 12:07

Simply way is to dissolve a weighed sample in water, put in a distilling flask, add excess NaOH or KOH solution, and distill the ammonia into a receiver containing a known amount of standardised acid (H2SO4 is most commonly used) and then titrating the excess acid.

A simpler variation is to put the same amount of a NaOH solution into 2 beakers, add the sample to one of them, and boil both side-by-side for awhile to drive off ammonia. Then titrate the remaining NaOH in each beaker and calculate the amount of NH3 from the difference in the two results. In your case the Mg(OH)2 precipitate will slow down the titration, as it must react as well.

http://tinyurl.com/67wjhw

There are fancier methods for determining small amounts of ammonia :

http://www.biolbull.org/cgi/reprint/67/1/126.pdf

matei - 16-9-2008 at 12:58

A very sensitive reagent for ammonia is Nessler's reagent -potassium tetraiodomercurate(II). It's original use was in qualitative tests, however a very simple spectrophotometric (or colorimetric - which is semiquantitative but useful if you don't have a spectrophotometer) method for ammonia can be developed.
The principle of the method is:
NH4+ + 2[HgI4]2− + 4OH− → HgO·Hg(NH2)I + 7I− + 3H2O

For small concentrations of ammonia, the reagent gives a yellow color. At higher concentrations, HgO·Hg(NH2)I precipitates.

I've checked ("ACS Reagent Chemicals") and for manganese chloride the ammonia quantity isn't determined. However, as an example, this is what they say about the ammonium assay in magnesium chloride:
Dissolve 1.0 g in 90 mL of water and add 10 mL of freshly boiled 10% sodium hydroxide reagent solution. Allow to settle, decant 50 mL and add 2 mL of Nessler reagent. Any color should not exceed that produced by 0.01 mg of ammonium ion (NH4) in an equal volume of solution containing 5 mL of 10% sodium hydroxide solution and 2 mL of Nessler reagent.

[Edited on 16-9-2008 by matei]

Attachment: cas7791_18_6.pdf (55kB)
This file has been downloaded 851 times

matei - 16-9-2008 at 13:13

For the preparation of Nessler's reagent:

Attachment: reagent_sol.pdf (97kB)
This file has been downloaded 1169 times

DJF90 - 16-9-2008 at 14:49

I dont know about others here but I tend to avoid mercury and its compounds due to environmental issues and difficulties in disposing of it properly. Is there not an equivalent reagent that is more "green" and easier to work with?

not_important - 16-9-2008 at 15:17

Quote:

Originally posted by DJF90
I dont know about others here but I tend to avoid mercury and its compounds due to environmental issues and difficulties in disposing of it properly. Is there not an equivalent reagent that is more "green" and easier to work with?

The 2nd reference I gave uses hypobromite and a dye, the end point being where excess OBr- reacts with the dyestuff.

DerAlte - 16-9-2008 at 21:45

@blogfast

Interpolating your intended usage from other posts, why are you so concerned about the NH4Cl? Can't you sublime it off by heating? Sublimation temp = 338C (CRC). MnCl2 won't even melt, let alone decompose, at this temp. And even if there's a little NH4Cl left, won't it just evaporate under your thermite conditions? Am I missing something?

@DJF90

Only a "green" would care a rat's ass about throwing about 70 mg or whatever of Hg compound down the john, flushing and saying goodbye. It a bit like that nonsense of the glob of mercury on that LA train station platform - call out the HAzmat boys! Paranoia reigns rife these days.

Matei makes a very good suggestion, If Blog is really that bothered about the ammonium level.

Der Alte

blogfast25 - 17-9-2008 at 07:44

Thanks gents!

Good old Nessler, completely forgot about him...

Quantitative testing (alkali + nose) showed in the past that there was residual NH4Cl.

Back titration standard H2SO4 into which the NH3 was distilled, is doable, even with my limited set up.

I also estimated the Mn content, expressed as MnCl2, of the said anhydrous MnCl2, be precipitating as MnCO3. That resulted in an estimate of > 100 w%, so despite obvious measuring error it's safe to assume the level of NH4Cl must be quite small.

Further glowing of the anhydrous MnCl2 at 400 C in dry CO2 and carefully weighing before and after could also give a good idea.

@ DerAlte:

The reason why I want to determine NH4Cl quite accurately is that I'm having problems with volatility in the MnCl2/Mg reductions, leading to highly porous slag/metal mixes and no separating out of the metal from the MgCl2.

The most likely cause is volatility of the MgCl2: with a BP of 1,412 C this will be quite volatile at the MP of Mn (1,246 C) with vapour pressure approaching 1 atm.

If this is the problem, then carrying out the reaction in a pressure resistant bomb-type reactor may be the only way to exceed the MP of Mn without blowing off the slag (never a good thing). In an open reactor, the latent heat of evaporation of the MgCl2 would also cost heat and cool the reaction down.

Before I resort to a pressure vessel, I need to be sure though that NH4Cl volatility isn't causing the problem, hence my question.

For a small pressure vessel, I'm thinking the cylinder and piston of a defunct RC plane engine...

DJF90 - 17-9-2008 at 09:32

i thought RC plane engines were made from aluminium?? Would this handle the temperatures you intend to throw at it (even under pressure...)

blogfast25 - 17-9-2008 at 09:59

DJF90:

The engine bloc, piston etc would act as a heat sink: the total amount of heat released, say ΔH, would cause the mass m of the assembly of engine + reaction products to heat by ΔT = ΔH / (C m), with C the average heat capacity of the assembly (engine + reaction products). If mass of the engine bloc is sufficiently high, its temperature will remain modest. In any case, the end-temperature of the assembly can be estimated roughly.

Superficial damage to the cylinder interior wall could be avoided by lining but wouldn't be that important in any case, as the engine would not be used for anything else anymore...

[Edited on 18-9-2008 by blogfast25]

Engine pistons

tapira1 - 21-9-2008 at 10:27

Unless the material they are made has changed, your engine pistons will melt before your salt will even notice the heat. You need a stainless steel bomb with a screw-cap

blogfast25 - 22-9-2008 at 08:26

Quote:

Originally posted by tapira1
Unless the material they are made has changed, your engine pistons will melt before your salt will even notice the heat. You need a stainless steel bomb with a screw-cap

... is a very sweeping statement that ignores many aspects, such as heat capacity, heat transfer rates, the actual amounts of heat involved (and not just the temperatures) and other factors.

While it's true that the reaction leads to temperatures in the region of 1,200 - 1,300 C, so well above the MP of most materials used in combustion engines, that simply doesn't imply the piston or cylinder wall will come even near that temperature.

For instance, a couple of days ago, I coated the bottom of a steel crucible with alumina by using the oxidation of Al powder with KClO3 (KClO3 + 2 Al ---> Al2O3 + KCl, ΔH @ 298 K about -1,255 kJ/mol of chlorate). The excessive amount of heat released ensures that the alumina (MP 2,054 C) forms in the molten state and thus can 'cast coat' the surface of the crucible. Well, the coating was quite successful, at least for a first "let's see" attempt and the crucible survived effortlessly.

But had I used more of the Al/chlorate mix on a much thinner crucible, the reaction would have ended up with a molten puddle of steel/alumina...

BTW, with a bit of boosting the reaction I'm studying can very easily surpass the MP of steel as well: even with a stainless steel bomb with a screw cap too large a charge could seriously damage, even destroy an SS bomb reactor. It's all about ratios... not absolutes.

watson.fawkes - 22-9-2008 at 16:28

Quote:

Originally posted by blogfast25
For instance, a couple of days ago, I coated the bottom of a steel crucible with alumina by [reaction elided]. The excessive amount of heat released ensures that the alumina (MP 2,054 C) forms in the molten state and thus can 'cast coat' the surface of the crucible. Well, the coating was quite successful, at least for a first "let's see" attempt and the crucible survived effortlessly.

But had I used more of the Al/chlorate mix on a much thinner crucible, the reaction would have ended up with a molten puddle of steel/alumina...

There's a lot to recommend this procedure as regards the lifetime of the result. Many metal-ceramic hybrids have trouble with thermal expansion. In this case, the formation temperature is hot, and this is the point of "neutral" internal compression and tension. The thermal coefficient of every common steel is greater than that of alumina. So when the steel/alumina hybrid shrinks to room temperature, the alumina is under compression (good for ceramics) and the steel under tension (just fine for metals). Thus as long as the surface bond is good enough and/or the internal curvature is "spherical" enough (a perfect sphere needs no surface bond, e.g.), the hybrid will in good shape.

Indeed, as long as the operating temperature is less than the temperature of manufacture (the melting point of alumina), the alumina will remain under internal compression. The acts just like a prestressed concrete beam, that is, it's far more resilient against shocks. In order to get the full benefit, though, the steel should be at the melting point of alumina to ensure that it's free of internal stresses at the point of solidification of the alumina.

blogfast25 - 23-9-2008 at 03:50

@ watson.fawkes:

Interestingly, it's my intention to try and develop this technique a little further by applying the coating to the crucible while heated to at least red heat. Probably by applying the unreacted chlorate/aluminium mix to the steel crucible and then furnace heating until the reaction occurs.

If successful, it would be a way to cheaply upgrade steel crucibles. The later have some interesting advantages to other materials such as ceramics or clay-graphite (but also the well-known drawbacks).

watson.fawkes - 23-9-2008 at 12:44

Quote:

Originally posted by blogfast25
Interestingly, it's my intention to try and develop this technique a little further by applying the coating to the crucible while heated to at least red heat. Probably by applying the unreacted chlorate/aluminium mix to the steel crucible and then furnace heating until the reaction occurs.

Consider, also, using your perchlorate/aluminum mixture as a reactive binder to sinter alumina powder. If it's hot enough, the outside of the alumina crystal will melt and then later fuse. You can tinker to get a good mixture of enthalpy release from the binder and heat capacity of the alumina.

I did have a question about the KCl left over. With its fairly low melting point, I'd imagine that it just remains liquid at the surface, but entrained droplets would be a source of pinhole defects in your original surface coating experiment. When sintering, it could be likewise a source of porosity. Did you inspect your coating for smoothness and defects?

Lastly, I'd imagine that a washing the steel with thin coat of sodium silicate would improve adhesion by formation of a layer of kaolin and jadeite (amongst other species) that will provide mechanical buffering and a more diffuse boundary interface.

blogfast25 - 24-9-2008 at 05:25

Quote:

Originally posted by watson.fawkes

I did have a question about the KCl left over. With its fairly low melting point, I'd imagine that it just remains liquid at the surface, but entrained droplets would be a source of pinhole defects in your original surface coating experiment. When sintering, it could be likewise a source of porosity. Did you inspect your coating for smoothness and defects?

My initial test was really a rather "dumb" let's-have-a-go type test and the obtained surface coating was rough and ready and much improvement can be envisaged. Some of the surface irregularities may well be due to evaporating KCl residue.

Some pretreatment of the steel surface, to prime it for subsequent coating is of course in order. Your suggestion of sodium silicate is noted. I was also thinking of an acid wash, to improve small scale surface cratering, much like is used to prime aluminium for treatment with molten Teflon.

Adding other ingredients to the chlorate/Al mix, such as alumina, kaolin or silica, is also a possibility, to impart desired properties to the coating. Not ethat I use chlorate, not perchlorate.

[Edited on 24-9-2008 by blogfast25]

chochu3 - 24-9-2008 at 16:05

Copper sulfate can be used for a quantitive test of ammonia and amino compounds not sure about ammonium compuonds, but worth a try.

blogfast25 - 25-9-2008 at 04:58

Quote:

Originally posted by chochu3
Copper sulfate can be used for a quantitive test of ammonia and amino compounds not sure about ammonium compuonds, but worth a try.

You mean via the deep blue Cu (II) - ammonia complex?

DJF90 - 25-9-2008 at 06:21

One of the little tests we did with amines at school was to add some copper (II) sulphate. We ended up getting a brown like complex. I'm sure it will vary depending on the ligand, which I'm afraid I can't remember what it was.

byko3y - 27-3-2016 at 15:41

I've spent some time looking over internet for the ammonia test. And I'm really confused by the fact that every book and article recommends Nessler reagent, despite the fact there's other methods. There's even an article in wikipedia about at least one alternative reagent Berthelot's reagent, which has selectivity ammonia-amines at least 10 to 1, at the same time being extremely easy in acquisition (phenol and hypohalite), much easier than o-(Benzenesulfonamido)-p-benzoquinone reagent, which is not so selective; while the distillation-hypobromite titration route is a tricky one to perform, takes a lot of time and devices.
I've just performed the berthlot's test to detect 20µg of ammonia in 5 ml solution, the coloration is so intensive I'd say I will readily detect 1 µg of ammonia in this solution. Looking at papers after writing the last sentence, I can see that actually 1/100'000 dillution of NH4Cl can be detected as a visible change in color, AFAIK which is comparable to sensitivity of Nessler's reagent.
My only question for now is: why the hell do everybody talk about nessler's reagent only? Is it some ideological inertia, like "we have a well know method we have been using for few hunder years, why shall we look for something else?". And even OTC stuff seems to be based on mercury halide reaction Subject: Mercury(II) iodide

chemrox - 27-3-2016 at 17:31

Quote: Originally posted by matei

For the preparation of Nessler's reagent:

Nice resource-thanks. For quantitative work I believe the Kiejdal method is usually used. One could also apply an ion specific electrode for ammonia as it is liberated via base. As long as you bias in some ammonia to get on the linear part of the curve the latter would be more sensitive to small quantities.

[Edited on 28-3-2016 by chemrox]

byko3y - 28-3-2016 at 07:21

Kjeldahl is about decomposition of nitrogen organic compounds into ammonia and then back-titration of unreacted acid (usually boric acid).
Ion-selective electrodes are expensive and complicated in usage, requires experienced chemists to be operated correctly and have limited lifespan.
Also, I don't think matei can read your message.