Sciencemadness Discussion Board

Reduction of Copper (II) Oxide with Ammonia from Decomposition of (NH4)2CO3

Bryce - 24-4-2017 at 14:28

Introduction / Motivations

Chemical methods for reducing copper ions to the metal can be very useful for making nonconductive surfaces conductive for electroplating. This is very important in printed circuit board (PCB) manufacture, in particular, for achieving reliable plated through-holes. Industrially, electroless copper plating is used, but this is often dependent on expensive reagents (e.g. palladium chloride) and proprietary bath formulations that may not be available to individuals in small quantities. For hobbyists and small companies, a process devised in post-Soviet Russia has been successfully adapted, reportedly achieving excellent results. Unfortunately, it depends upon the pyrolysis of a hypophosphite salt. Such salts are difficult to obtain in the USA due to the ineffectual flailing of the government in confronting the methamphetamine epidemic as a problem of chemical availability rather than social conditions. It is therefore desirable to develop an alternative electroless copper plating system which does not depend upon restricted chemicals or expensive proprietary plating baths.

Concept

Copper (II) oxide (CuO, cupric oxide) can readily be reduced to the metal, with moderate heating, by hydrogen, carbon monoxide, or methane gas, e.g:

CuO (s) + H2 (g) ---(heat)--> Cu (s) + H2O

Usually, the reaction is performed in a glass tube, heated by a burner, and containing a mound of CuO powder. The reducing gas is blown over the material slowly as it is heated and gradually reduced, the water and excess reducing gas escaping out the tube to the atmosphere. I hoped to create, instead, a mixture that could be spread on an irregularly shaped object, e.g. a PCB, and then heated in an oven to complete the reaction, leaving the object coated with copper metal.

However, hydrogen and methane are inconvenient to generate in situ. Hydrogen is typically generated from the action of acids on metals, e.g:

2HCl (aq) + 2Zn (s) ---> H2(g) + 2Zn+ (aq) + 2Cl- (aq)

Methane can be generated from water using aluminum carbide:

Al4C3 (s) + 12 H2O (l) ---> 3 CH4 (g) + 4 Al(OH)3

However even if physical circumstances could be arranged, the byproducts of these gas-forming reactions (e.g. zinc chloride) are undesirable to leave on the surface of the plated object. They would require some kind of special heated, controlled atmosphere apparatus and a separate gas generator (or bottled gas.)
This implies a considerably greater expense for anyone desiring to use the technique, in comparison to the copper hypophosphite pyrolysis method. Carbon monoxide was discounted because its insidious toxicity renders it inappropriate for widespread use by non-chemists.

At this point, I hit on the idea of using ammonia gas (NH3) to reduce CuO. A preliminary search online suggested this might be possible. Ammonia, unlike the other gases mentioned, can be generated in situ by the thermal decomposition, at very moderate temperatures, of ammonium carbonate:

(NH4)2CO3 --(120°C)--> NH3 + CO2 + H2O

All three of the products are gases at the temperature of decomposition, so no inconvenient residue remains. Furthermore, ammonium carbonate is promiscuously available as a chemical leavening for baking, and generally considered quite safe. Ammonia gas is, of course, poisonous, but unlike e.g. carbon monoxide, an increasing concentration of ammonia makes its presence known by its sharp, unpleasant odor long before reaching levels that pose an immediate danger to health.

I proceeded to investigate the reduction of copper (II) oxide with ammonia generated in situ by the decomposition of (NH4)2CO3.

Experimental

First, I prepared a mixture of CuO and (NH4)2CO3 for the hypothetical reduction of Cu+2 to Cu0:

* (NH4)2CO3 + 3CuO --(heat)--> 3Cu (s) + N2 (g) + CO2 (g) + 4H2O (g)

This hypothetical reaction taking place in two parts - the thermal decomposition of the ammonium carbonate followed by the reduction of the copper (ii) oxide by the ammonia released from the decomposition.

(A) I weighed out 1.0968g of CuO and 0.5657g of (NH4)2CO3 (versus the stoichiometric amount of 0.4416, a slight excess.)

A tiny amount (about 160 mg) was first burned unconfined to verify the absence of any unexpected energetic properties. There was nothing untoward. Residue had a reddish tinge.

Next, 393.9 mg were placed in a 10x100mm test tube and heated in a gas flame. On heating, white fumes were evolved and condensed on the walls of the tube. Water also condensed. Vacuoles were seen to form in the powder. A red deposit formed at the bottom of the tube. I accidentally spilled it while cooling, and was unable to weigh the residue. The gases evolved did not appear to be flammable, but ammonia has a narrow range of flammability in air, to say nothing of a mixture of ammonia gas with five moles of non-flammable gases.

Observing that reduction seemed confined to the bottom of the tube, I thought confinement of the generated ammonia gas might be a problem. About 200 mg of the mixture was placed in an aluminium foil packet, which was heated to orange heat. Discoloration was found inside, but no convincing signs of reduction.

The remaining mixture was burnt in an open flame. White bubbling globs, probably (NH4)2CO3 decomposing, appeared, with a crust of red material consistent with Cu2O, or possibly amorphous Cu metal.

(B) Believing that some factor, perhaps physical factors or the difference in temperature between decomposition of (NH4)2CO3 and the reduction of CuO by NH3, I decided to use a considerable excess of (NH4)2CO3 in subsequent experiments.

0.4026g of CuO were mixed with 1.7119g of (NH4)2CO3 and heated in a test tube. The results were similar to the first experiment in a test tube, but with a noticeably greater quantity of reduced material appearing. The majority of the CuO did not seem to be changed, however.

290 mg of the second mixture was placed in an Al foil packet and heated to orange-yellow heat. A modest amount of the red, presumptive reduced material was found, along with similar discoloration to the foil.

A second experiment was performed with a test tube. 0.9892g of the mixture were heated. The yield appeared to be modest, and I think I agitated it too much out of fear of a gas pocket developing in the larger amount of material. 0.1167g of material remained, not uniformly reduced (i.e. some was still black.)

The experiment was repeated with less agitation. 0.5019g of mixture was heated, but it was well-adhered to the tube and I accidentally spilled the weighing boat. The yield appeared to be considerably greater than the ~100mg of the previous run, though, despite the lower quantity of starting material.

(C) At this point I switched to filling the bottom 11mm or so of the test tube with (NH4)2CO3 (without weighing it exactly) and putting carefully-weighed amounts of CuO on top of it, without much mixing. The idea here was that the tube would be heated with the flame under the CuO, ensuring it would already be quite hot when conducted heat began to decompose the (NH4)2CO3 and liberate ammonia. This appears to have greatly improved the results, subsequent experiments yielded what appeared to be entirely reduced material, though no analysis was done to confirm the absence of Cu2+ quantitatively.

The experiment with the layered materials was repeated 3 times, with the following results:
1. 112.0mg of CuO -> 81.1mg of red material.
2. 136.5mg of CuO -> 102.7mg of red material.
3. 319.3mg of CuO -> 174.9mg of red material.

On run 3, black smoke was twice seen, associated with agitation. A greater excess of (NH4)2CO3 had been used. I conjecture that the "smoke" was CuO propelled out of the tube by escaping gases, reducing the yield somewhat.

Analysis of the reduced material: The reduced material was red to slightly orange. It was found to be slightly conductive of electricity. Strong ammonia turned pale blue when it contacted the reduced material, eventually becoming deep blue. A drop of 70% nitric acid on the reduced product in air produced orange gas vigorously, resulting in a sky blue solution. The reduced product sintered but did not fuse in a MAP-substitute gas / air flame, and turned black in an oxidizing flame. It gave an unimpressive green flame test.

Analysis

The reduction of CuO by ammonia is energetically favorable. I believe what happens is, in the presence of excess ammonia:

6 CuO + 2NH3 --> 3Cu2O + 3H2O + N2 + 339.4 kJ

3 Cu2O + 2NH3 --> 6 Cu + 3 H2O + N2 + 255 kJ

The identity of the reduced material is not completely clear yet. Unfortunately, as a red powder, its appearance is consistent with Cu2O (cuprous oxide) or Cu metal powder. I thought the nitric acid test would be helpful, but then I read that Cu2O also gives the same result as copper.

The amounts of reduced material, assuming pure Cu2O, obtained in the layered tube experiments (C) correspond to 0.56 mmol, 0.71mmol, and 1.2 mmol respectively. On the other hand, assuming pure Cu, they would be 1.3 mmol, 1.6 mmol, and 2.8 mmol. These came from 1.4 mmol, 1.7 mmol and 4.0 mmol of CuO. (The last tube was the one that bumped black smoke.) This suggests it was copper, but if so, its electrical conductivity was rather less than was hoped, at least as a slightly compressed powder. A suggestion for a simple qualitative test that doesn't require a titration would be welcome. If I don't get one, I plan to, time permitting, figure out how much copper per gram of reduced material I have by an appropriate titration, and thereby determine if it is Cu or copper (i) oxide.

However, in any case I still hope this might be useful for someone who e.g. wants to explore reductive chemistry using a common baking ingredient and readily available copper (ii) compounds. The reduction of other copper (ii) compounds by ammonia gas generated in situ with the layered test tube technique described above might be an interesting direction; indeed, perhaps it could be applied to other chemistry involving small quantities of gaseous ammonia. (Providing they can tolerate the presence of water and carbon dioxide.)


[Edited on 24-4-2017 by Bryce]

[Edited on 25-4-2017 by Bryce]

ninhydric1 - 24-4-2017 at 14:50

On the Wikipedia page for copper(I) oxide, it says that copper(I) oxide will dissolve in HCl to form CuCl2- ions. Since HCl doesn't dissolve metallic copper, you could use HCl to remove the copper(I) oxide from the red material you obtained from the experiment.

https://en.wikipedia.org/wiki/Copper(I)_oxide

Praxichys - 24-4-2017 at 15:02

Mix some of the red material with powdered aluminum and hit it with a torch. If it's Cu2O, you'll get a thermitic reaction. If it's Cu metal, there won't be a reaction.

Bryce - 24-4-2017 at 16:22

Things I did not learn in college chemistry: attempting to making thermite out of the unknown is a cornerstone of practical qualitative analysis :D

I performed both tests suggested - thank you for the ideas.

Results:
I first prepared a fresh batch of reduced material with the layered test tube method described above. 0.3387g of CuO (4.258 mmol) resulted in 0.2598g of reduced material (1.816 mmol if Cu2O, 4.088 mmol if Cu) of which 0.2542g was put into concentrated HCl. There was no fizzing and no discoloration of the HCl even after mixing and standing a bit. The solid material remaining in the flask (apparently all of it, as far as I could tell) was vacuum filtered and rinsed with water, then dried on a hot plate. It looked the same as it did when it went in to the acid. I was unable to weigh it due to complications with the drying.

I mixed 0.1749g of the recovered reduced material with 40.5 mg Al powder (stoichometric amount if the reduced material were Cu2O: 39.6mg). I noted that it is a pain in the butt to try to make 0.18 grams of thermite. The mixture was exposed to the flame of a torch. Some copper-color in the flame was seen, though not to an impressive degree. There was no thermite-type reaction seen. The surface of the material turned black. Despite the surface at least having reached yellow heat in the flame of the torch, the interior, when the surface was broken, remained like the original Al/reduced material mix.

On this basis, and that of the further confirmation of the gravimetric results (the mass of the product corresponds closely with Cu and poorly with Cu2O), I conclude that ammonia from the decomposition of ammonium carbonate can reduce copper (ii) oxide.

Thanks again for the suggested tests.

[Edited on 25-4-2017 by Bryce]

AJKOER - 26-4-2017 at 12:28

Here is a link to an experiment involving hot CuO and ammonia: https://www.yaaka.cc/unit/ammonia/ . Note, dry NH3 is cited as an input, with the formation of brown-reddish copper metal. Your embodiment, employs NH3/water vapor, and is a departure from procedure.

In general, do not forget the equilibrium reaction that occurs in the presence of copper metal and cupric forming cuprous:

Cu + Cu(ll) = 2 Cu(l)

If you allow, or at the end if the experiment, let oxygen into the system:

--- With hot copper and oxygen, NH3 can be converted into NO (also NO2 depending on O2 concentration) and water vapor.

-- Ammonia complexes with copper (in the presence of water and oxygen) forming, for example, tetraammine Copper(ll) (or Copper(l) ) hydroxide, and, in the presence of CO2, possibly corresponding carbonate compounds. There is actually nothing sacred about the tetraamminediaquacopper cation as depending on the ammonia concentration in H2O, one could have anywhere from [Cu(NH3)(H2O)5]2+ to [Cu(NH3)5(H2O)]2+ with varying changes in color intensity. In any event, under further thermal decomposition the product is either CuO or Cu2O, NH3 and water vapor.

-- Also in the presence of any oxygen, there could also be an electrochemical reaction occurring as well (see reference below). In the system of Cu/NH3/H2O/O2 there is also a minor pathway (in essence, a standard chemical reaction) that results in some NH4NO2 via a REDOX reaction. The latter is note worthy as ammonium nitrite generally safely decomposes liberating a good quantity of nitrogen gas.

Reference: https://www.academia.edu/292096/Kinetics_and_Mechanism_of_Co...

[Edited on 27-4-2017 by AJKOER]