CR Scientific

Newsletter Archive - Favorite Articles

You may notice that the end-notes don't necessarily start at "1".  This article is an excerpt from a back issue of our newsletter.

Uranophane Experiment -
Ammonia Precipitation
by C. Thorsten

Analysis of Uranophane using Ammonia Precipitation
Safety reminder:  Uranium salts are toxic.  exercise due caution when handling radioactive materials.  Do not ingest or inhale the particles of uranium compounds.
Uranium's primary hazard is not its radiation (which is mild), but rather its toxicity.  It is similar to the other "heavy metals" (lead, nickel, chromium) in this respect.  Grinding up a uranium mineral for fusion presents an inhalation hazard.  You must wear a cartridge respirator that can filter out harmful dusts.   For safety, assume that soluble heavy-metal compounds will also be absorbed through your skin;  exercise appropriate precautions.  
It is beneficial to have and use a Geiger counter to check your work area.  A wet cloth is good for picking up dust.  When trying to do a fusion on the carbon block, it's best (1.) to do the procedure outdoors, and (2.) to have a shallow hole drilled into the block where the mixture can be placed while hitting it with the torch flame;  this helps prevent the powder from blowing away.
Be careful not to boil uranium solutions too rapidly, as spattering may carry some of the dissolved uranium salts into aerosol form.
One of our readers emailed a summary of an experiment he'd done using a sample of uranophane from Canada.  The general procedure was from the first edition of Orsino Smith's Identification and Qualitative Analysis of Minerals (1946).  The sample was crushed and fused with sodium carbonate, and then the fusion was dissolved in distilled water.  To the resulting solution was added some solid ammonium chloride and then aqueous ammonia (i.e., ammonium hydroxide, NH4OH).  The solution was boiled and allowed to cool.   The filtrate from this ammonia step was then treated with ammonium oxalate and filtered;  the filtrate from the oxalate step was in turn treated with ammonium phosphate.
In this article we'll explore this procedure in some detail, especially the first step (ammonia precipitation).  To follow along with our reader's experiment, we performed the same procedure using a small bit of uranophane.
As we'll discuss later, a thorough separation of uranium from other elements can be a very involved process.   In the current experiment we'll look at the chemistry of a highly simplified procedure.

1.)  Crushing of the sample and mixing with sodium carbonate:   This was accomplished by breaking off a pinky-nail sized piece of uranophane from a rock sample and pulverizing it in a steel mortar and pestle.  Approximately three times the volume of powdered sodium carbonate (relative to the sample size) was added to this.

2.)  Fusion of the sample:   The uranophane-sodium carbonate mixture was placed in a shallow hole that had been drilled in a carbon block.  The mixture was fused with a propane torch until it melted into a single bead.

3.)  Dissolution of the sample:  The bead was allowed to cool and placed in 5 mL of distilled water in a micro beaker.  The block was checked for remnants using a Geiger counter;  considerable residue was found in this way.  It was scraped off carefully and added to the beaker. 
The bead and other residue were allowed to stand in the water overnight.

Figure 7.  The sodium carbonate fusion was dissolved in water.  Notice the very slight coloration of the water solution;  note also, however, the considerable amount of insoluble matter at the bottom.  Some of the black came from carbon particles scraped from the block.

When this precipitate was left behind on the filter paper, it was checked with a Geiger counter and found to contain most of the detectable radioactivity of the sample.  

The next day, the remainder of the water-logged bead was broken apart with the end of a glass rod.  The solution was stirred and allowed to settle again for a photograph (figure 7).  It was then stirred up and filtered via gravity.
This filtrate (which we'll call the "water fraction") was collected in a micro beaker and set aside.  It had only a very slight yellowish tint, barely noticeable against a white background.

3a.)  Insoluble residue remained in the filter paper.  This was checked with a Geiger counter and found to contain most of the sample's radioactivity.  This residue was treated with 6 M HCl;  the resulting filtrate (which we'll call the "HCl fraction") was collected in another micro beaker (figure 8). 
A drop of this solution crystallized on a microscope slide yielded distinct crystals of NaCl, but only shapeless, vaguely-crystalline yellow masses of uranium salts at this point.

Figure 8. The filtrate from step 3 did not contain much uranium, judging from its radioactivity and its color.
Most of the uranium was in the "HCl fraction", the solution obtained by dissolving the Step 3 precipitate with hydrochloric acid.

The HCl solution remained transparent yellowish but acquired a gelatinous consistency on standing.  This was presumed to be silicic acid.

4.)  [Ammonium Chloride and] Aqueous Ammonia:   As discovered in step 3, the bead contained water-insoluble matter that had to be carried into solution with HCl.  Because this introduced chloride ions into solution and thus formed NH4Cl, step 4 omitted the ammonium chloride.  
About 3 mL of the HCl fraction were treated with 10% aqueous ammonia (presumed to contain some dissolved CO2).  At first the solution went from pale yellow to colorless.  More ammonia was added until the solution had a slightly ammoniacal odor.  At this point the solution turned yellow again.  On standing a muddy yellow, cloudy to slightly gelatinous precipitate began to form (figure 9).

Figure 9.  The solution from Step 4.  Notice the muddy-yellow, flocculent  precipitate.  At this point, the solution was presumed to contain NH4Cl as well as ammonia, since the aqueous NH3 had been added to a solution containing HCl.

The photo shows the solution before boiling.

5.) Boiling:   The micro beaker containing the ammoniated solution was placed on a wire gauze on a ringstand.  Five, PTFE boiling chips were added (to prevent "bumping").  The liquid was heated from underneath with an alcohol burner.  The precipitate darkened upon heating.   Boiling was continued for 5 minutes.  The heat was removed and the beaker allowed to cool.  A bright yellow, somewhat flocculent precipitate settled to the bottom (figure 10).

Figure 10.  Now we're getting somewhere:  the solution after Step 5.  The Teflon chips are still in there.  
Note the color change to bright canary yellow. 
The yellow is probably UO22H2O. 
In our sample the hydrous U oxide is mixed with silica, silicic acid, insoluble Na silicates, etc. 

6.) Filtration:   The solution was poured through filter paper in a funnel.  The bright yellow precipitates on the funnel were washed with distilled water.  The washings were collected in the filtrate.  This solution was saved for Step 7.  The yellow precipitates themselves were checked with the Geiger counter;  predictably, they were radioactive.  These precipitates were saved for Step 10.

Figure 11. The precipitate from Step 6:  bright yellow
and detectably radioactive. 
The precipitate is
made bulkier by the presence of silica and / or

silicic acid, which are separated in Step 10.

7.)  Ammonium Oxalate:   A 10 x 75 mm test tube full of the filtrate from step 6 was treated with saturated  (NH4)2C2O4.  This solution was allowed to stand for a few hours.   A very fine, diffuse, white precipitate slowly formed, indicating Ca2+ / Mg2+ / Sr2+.  However, there wasn't much of it, since the Step 6 filtrate had been diluted to a volume of 60 mL during the washings.
Boiling down the Step 6 filtrate to 15 mL and trying again produced a definite, white precipitate on treatment with ammonium oxalate.  Allowing the test tube to stand for a week produced fairly large, colorless crystals.  The crystals appeared to be finished growing after two weeks.

Figure 12. Oxalate precipitate from Step 7. The test
tube was allowed to stand for two weeks to allow
complete precipitation of insoluble oxalates.  These
crystals took quite a long time to grow to the stage
shown here.

Click for full-size image.

8.)  Filtration to remove precipitates from step 7.  The precipitates had actually settled completely after two weeks;  the liquid was siphoned off with a dropper.

9.)  Ammonium Phosphate:   The liquid from step 8 was evaporated down to about 1/4 its volume by gentle heating, allowed to cool, and treated with (NH4)2HPO4 (which can be made by treating phosphoric acid with two molar equivalents of NH4OH).
The solution was allowed to stand for a few hours.

10.)  Treatment of the Precipitate from Step 6:  The water-washed precipitate on the filter was dissolved by running some 6 M HCl through the filter paper.  The yellow filtrate was collected in a clean beaker.  Left behind was a clear, gelatinous precipitate on the filter.

Figure 14.  Silica / silicic acid remaining on the filter after
dissolving the uranium precipitates with 6M HCl. 
If this had been Al(OH)3 or other metal hydroxide,
it would have dissolved in the strong acid.

Figure 15.  The solution from Step 10, made by dissolving the bright yellow uranium precipitates in 6M HCl and washing with water, followed by more HCl.

This solution at this stage is strongly acidic.  It will contain some NH4Cl unless the precipitate at Step 6 was washed very thoroughly and perhaps boiled in distilled water to remove adsorbed ammonium compounds. 

Some of the Step 10 solution was placed on a microscope slide and allowed to evaporate at room temperature.  Square, bright-yellow crystals formed;  these were thought to be ammonium uranyl chloride.  There were no appreciable amounts of other crystal species present in several of these slides. 

10a.)  A sample of this Step 10 solution was placed in a crucible, evaporated to dryness with an alcohol lamp, and strongly heated for at least 5 minutes to drive off any ammonium salts.  The residue, when cool, was redissolved in 7 M HCl.  This new solution was also yellow, but this time it was impossible grow the beautiful, kaleidoscopic yellow squares. 

Figure 16.  A crystal of ammonium uranyl chloride, viewed at 400x.  These crystals displayed interesting and often remarkably beautiful patterns within their square outlines.  They were reminiscent of quilt patches or kaleidoscope patterns.

Click for full-size image.


We've explored a highly simplified separation procedure for uranium which, in the specific case of uranophane as the starting material, worked well enough for qualitative purposes.
During the carbonate fusion, the uranophane sample was converted to a mixture of the corresponding metal hydroxides and oxides;  the fusion also contained sodium hydroxide (from the decomposition of the carbonate),  sodium silicates, and some undecomposed sodium carbonate.
Treatment of the bead with water solubilized the Na, some of the Ca, and the greater part of any Al and Pb that might have been present.  The bulk of the soluble silicates of Na were also dissolved at this step;  these gradually separated as insoluble silica on standing for a few days.  Some of the silicates, however, did remain undissolved in the water at step 3;  these found their way  into HCl solution during Step 3a.
The HCl-soluble portion of the bead (Step 3a) contained the U, the rest of the Ca, and any Mg, as well as any other elements whose oxides or hydroxides weren't soluble in alkaline solution.  The carbon-block residue in Step 3 was presumed to contain uranium oxides, which would have formed during intense roasting.  This would explain their insolubility in the alkaline solution resulting from the "water fraction" (Step 3).
Treatment of the HCl fraction with excess ammonia (Step 4) introduced NH4Cl into the solution, as well as aqueous NH3.  The NH4Cl kept any Mg2+ in solution while the U2+ precipitated at Step 5.  It's important to note here that the uranium wouldn't have come down if there had been appreciable ammonium carbonate or vanadium in solution.   (We can safely say, therefore, that the specimen was indeed uranophane, not carnotite;  the latter contains much V).
Boiling the ammoniated HCl fraction (Step 5) caused formation of hydrous uranium oxide, as previously mentioned.  Step 5 also brought about decomposition of any alkali silicates that had dissolved in the HCl fraction.  The mixture of hydrous uranium oxide and silica gel / silicic acid stayed together until Step 10, when the siliceous compounds were left behind on the filter paper and the uranium compounds were carried into solution by HCl.  At Step 10 it appeared that uranium was the predominant metal cation in solution.  However, the author suspected that the yellow crystals were actually a double salt of ammonium and uranyl ion.  Some residual NH4Cl must have remained adsorbed to the UO22H2O and silicic acid precipitates from Step 5, continuing to adhere through the washings of Step 6.  Evaporating some of Step 10 solution to dryness, heating intensely enough to drive off NH4 compounds, and re-dissolving the residue in HCl gave a solution from which no yellow crystals could be grown.  At the temperatures used, Na compounds would not have volatilized. 

Figure 17.  The absence of other crystal types suggested a
lack of appreciable impurities by Step 10, although the
square, yellow crystals were probably a double salt with
 ammonium chloride.  Heating the solution to dryness
and volatilizing the NH4 salts, then redissolving in HCl,
gave a solution of what was expected to be uranyl chloride. 
The putative uranyl chloride solution did not yield crystals on any attempt.

It seems the solution from Step 10 had just the right molar
proportions of ammonium chloride and uranyl ion to grow
these crystals.  This was an accidental but pleasing
outcome of the experiment.

Ammonia precipitation is a useful technique, but by itself it acts on perhaps too wide a spectrum of ions unless one takes care to adjust pH and other conditions.  Assuming no other precipitations have been done first, aqueous ammonia can in theory precipitate ions of Fe, U, Cr, Hg, Bi, Ti, Zr, Th, Al, Be, Sn, Pb, Co, Cu, Zn, and Sb, mostly in the form of hydroxides and/or hydrated oxides.  The boiling, which is necessary to get all the uranium precipitated, is also what brings down so many other ions.  For example, Cu2+ in ammonia will at first stay in solution as the tetramminecopper (II) ion, but heating decomposes this and precipitates CuO.  An analogous situation can happen with Ag, Ni, Co, Cd, Zn, Cr, and even Fe.  Ammonia precipitation may also bring down some, but not all, of the Mn, Mg, and even a few others. 
The sheer number of ions that can be brought down by ammonia would seem too great to render the operation very useful for general work, at least when it is the first major treatment conducted on the sample.  The chemists who wrote qualitative analysis textbooks evidently thought so, too.  Their answer was to attempt to narrow the range of elements still in solution by the time the experimenter made it to the ammonia precipitation.  At least one, fairly standard routine has emerged from this.  Sorum's text (1953) uses it, but many other manuals this author has seen appear to contain the same basic scheme.
The first step in a standard qualitative analysis routine is precipitation of  the "silver group" using cold, dilute HCl.  Silver (I), mercury (I), and lead (II) should come down if present.  In our case, where we had a mineral suspected to be uranophane, we assumed there wasn't much of any of these.  Our assumption might have been wrong, especially in the case of lead, but we skipped the "silver group" precipitation anyway.  (It is worth noting here that no precipitate of PbCl2 was noticed in the HCl fraction from our experiment, although the amount of Pb2+ may have been small enough that the chloride remained dissolved;  it would then have been held in solution in the presence of NH4Cl during Steps 4 and 5.  Assuming, on the other hand, that Pb never made it into solution from the filter paper in Step 3a, we might have wished to soak the paper in aqueous NH4Cl and then test this solution for lead, perhaps using cyanidin chloride).
Next in a standard qualitative analysis scheme is to treat the acidified filtrate or decantate with hydrogen sulfide (danger- highly toxic) to get even more elements out of the way before going on to the ammonia step.  The so-called "hydrogen sulfide group" or "copper-arsenic group" will come down with acidified H2S:  tin, copper, arsenic, antimony, bismuth, cadmium, and the remaining lead and mercury.  Again, in our sample there probably weren't any of these elements present, so we were also able to skip the "hydrogen sulfide group" step. As stated before, however, one of the stable decay products of uranium-238 is lead-206;   it's therefore hard to imagine a natural uranium sample being entirely free of lead. 1  
Finally, when the qualitative analytical separation actually arrives at the ammonia step, it often combines the NH4Cl and NH4OH with ammonium sulfide.  This precipitates the sulfides of manganese, zinc, nickel, and ferrous iron, the bulk of which should not have come down with H2S treatment.  Furthermore, ammonium hydroxide / sulfide treatment will precipitate the hydroxides of aluminum, chromium (III), and ferric iron. 
In the abbreviated scheme used in the present experiment, however, we simply used aqueous ammonia and skipped a number of steps.  Though it wouldn't have been specific enough if a spectrum of metal ions had been present, we weren't expecting the silver group, the copper-nickel group, or the ammonium sulfide group in the sample.  Because the only metal ions we expected to encounter were those of uranium and calcium, we had some basis for our omissions. 2  
Even if a specimen definitely is uranophane (in this experiment, it was) and therefore doesn't have most of the ions that would occur in a qualitative analysis scheme, the fact that a few, similar-looking uranium minerals do contain Al, V, and / or Mg should have some bearing on the analysis.  After all, that mineral thought to be "uranophane" could easily be tyuyamunite, carnotite, or one of several others.  Therefore it is wise for us to consider a few additional points, even though we used a heavily simplified procedure:
  • Al(OH)3 and certain other precipitates will come down at first from ammonia treatment, but they'll redissolve in part if too much ammonia is added (excess NaOH redissolves Al(OH)3 completely).  This will cause noticeable traces of Al to carry into subsequent steps.  According to Hillebrand and Lundell (1929), the pH should be adjusted to about 7.0 to bring down Al as completely as possible.  It should stay below 7.5 to prevent re-dissolution.  In other words, don't add so much ammonia that the solution becomes noticeably alkaline.  If one wants to re-dissolve the hydroxides of Al, Zn, Cr, Sb, Sn, and Pb on purpose,  NaOH is more effective than NH3;  the latter doesn't form as much OH- in solution.  (One can then attempt to separate sodium aluminate, sodium zincate, etc. from these solutions;  see Sorum, 1953).  Uranophane doesn't contain aluminum in its formula, but a number of related minerals do.  It is probable that the matrix rock contains at least some Al3+.
  • Uranium ions will not precipitate completely if ammonium carbonate is present in the aqueous ammonia, because ammonium uranyl carbonate forms.  Enough ammonium carbonate will therefore keep all U completely in solution.  In our experiment there may have been minor amounts of ammonium carbonate already in solution, since we started with a sodium carbonate fusion and then had it in solution with ammonia (even though most of the carbonate was driven off by the intense heat of fusion, some was probably reabsorbed from the air on standing).  One way to get around this would be to roast the fusion very intensely in a crucible to drive off the CO2 (leaving hydroxides and oxides) then let it cool in a CO2-free atmosphere or a vacuum desiccator having soda-lime in the bottom.   Taking up this cooled fusion with freshly-prepared aqueous ammonia  (made with degassed, deionized water and pure NH3) would minimize the amount of ammonium carbonate in the resulting solution.  For qualitative work, though, a brand-new bottle of NH4OH is sufficient.
  • Vanadium ions will also tend to keep uranium ions in solution, causing them to carry over into the next step. (This can happen if, instead of "uranophane", the sample is actually carnotite)
  • Certain ions will co-precipitate species that wouldn't normally come out of solution.  For example, the ammonia will cause calcium phosphate to come down from solution if the mineral sample had any phosphate in it.  (This can happen if the mineral is autunite instead of uranophane)
  • Magnesium ions would precipitate during the ammonia step too, if it weren't for the ammonium chloride that Smith recommends adding to the solution prior to adding the ammonia.  A person is not born knowing such things.
With the complexities of ammonia precipitation, it seems hard to be sure if anything in our experiment either stays in solution entirely or precipitates  completely.  Since the focus here is qualitative rather than quantitative analysis, however, it is not as critical if a small amount of some ion carries into subsequent steps. 

If, by the way, we're sure a sample contains uranium-- for example, it's both yellow and radioactive-- but we get no yellow precipitate at the ammonia step, then we know that sample either contains vanadium with the uranium, or else we must have ammonium carbonate in the solution.  Perhaps both are present.  Chances are that we will have some carbonate left over from the fusion unless we roast the sodium carbonate to the point where it decomposes  to NaOH (somewhere about 1000-1200C).  Even then it will reabsorb significant amounts of atmospheric CO2 as it cools.   Unless the cooling happens in an atmosphere free of carbon dioxide, the NaOH will contain much Na2CO3; when this is dissolved in aqueous ammonia, there will consequently be the requisite ammonium and carbonate ions to hold uranium in solution.  There is also the matter of CO2 absorbtion by the ammonia solution itself;  a flexible plastic bottle of aqueous NH3 that has sat on a shelf for five or ten years is probably going to have absorbed some CO2.  Again, this can cause ammonia precipitation of uranium ions to be incomplete.

After the ammonia precipitation step, Smith's procedure calls for treatment of the filtrate with ammonium oxalate and then ammonium phosphate.  Let's examine why.
Not counting Na+ and K+, which would have entered the sample in huge amounts by way of the fusion anyway,  Smith's NH4OH / NH4Cl  treatment should hold Ca++, Mg++, and Ba++ in solution-- so long as there is no co-precipitation and we're fairly sure there aren't unaccounted-for metal ions in solution.   Of these three remaining species, uranophane contains only Ca++ in appreciable amounts.
The calcium ions should come down as insoluble calcium oxalate when the ammonium oxalate is added.  Oxalate precipitations become complete only slowly;  unless the solution is boiled for a few minutes, it can take days to  reach completeness.  Normally, most of the Ba++ and some of the Mg++ would also come down;  however, according to procedure P-19 in the second edition of Smith (1953), making the filtrate strongly alkaline with an excess of NH4OH before adding the ammonium oxalate should keep most of the Mg++ in solution, provided it's kept in the cold.  The oxalates of Ba and Ca, along with any Sr that might be present, should therefore be the only ones to come down appreciably.  If we used uranophane as our mineral sample, we might expect the oxalate step to be the last.
If multiple Group II ions are present in the initial sample, the only one that should remain in solution by the last stage is Mg++.  The final, phosphate treatment of the filtrate should thus bring this down as the insoluble magnesium ammonium phosphate (struvite), MgNH4PO4 6H2O.   A sample of this white precipitate (which we don't expect to have if our sample is pure uranophane and we do the procedure correctly) should produce a pinkish color when moistened with cobalt nitrate solution and heated on the carbon block (Smith, 1953).  
The mineral sample in this experiment behaved mostly as expected with regard to the various tests.  It is interesting that a failure to wash out all the NH4Cl at Step 6 created a useful result.  It will make a good future project to try duplicating the reagent concentrations necessary to allow crystallization of ammonium uranyl chloride.
For any future experiments involving uranium, we remind the reader once again to exercise caution.  Again, uranium's primary hazard is one of heavy metal toxicity;  wear a double layer of vinyl or latex gloves when handling soluble uranium compounds, and of course avoid breathing in dusts that may contain uranium.

- "Ammonia Precipitation"

1 As with the silver group, we may have to revisit this step if it appears our mineral sample contained more than we expected.
Back to the article

2 Impatient lab partners of the world, rejoice.  A bit of a digression here:  a number of students, even at the graduate level, make it a point to work as rapidly as they can, cutting corners as much as possible in an all-but-overt race to see who can "just get out of there" the fastest.  Some of these unfortunately go on to design or prescribe the medicines people take every day.  A few go on to handle surgical scalpels.   The analytical mindset, especially when applied to any kind of work where someone's life or future depends on the outcome, should have no room for the kind of habitual impatience this writer has seen on so many occasions. 
Editorializing aside, it's still not very thorough or scientific to cut out whole chunks of the qualitative analysis scheme if we want to allow that our mineral might not be uranophane at all.  Time permitting, we'd want to look for every common metal ion, at least the ones expected to occur in yellow, radioactive minerals that could be mistaken for uranophane.     
Back to the article

References - "Ammonia Precipitation"

Hillebrand, W., and Lundell, G. Applied Inorganic Analysis. New York: John Wiley and Sons, 1929.

Smith, Orsino C. Identification and Qualitative Analysis of Minerals, 1st ed.  Princeton: Van Nostrand, 1946.

Smith, Orsino C. Identification and Qualitative Analysis of Minerals, 2nd ed.  Princeton: Van Nostrand, 1953.

Sorum, C.H. Introduction to Semimicro Qualitative Analysis. New York: Prentice Hall, 1953.

Back to Articles Index

CR Scientific Catalog
Newsletter Index Page
Lab Glassware
Main Page