This experiment sets out to explore the possibility and feasibility of using liquefied gasses, in this case ammonia, as the reaction media for an experiment as opposed to the aqueous solutions where water is used. This experiment makes use of the fact that alkali metals, such as Sodium, are readily dissolved by liquid ammonia, but are highly reactive with water. Alkali metals dissolved in ammonia form blue coloured solutions in low concentrations and bronze when concentrated. The blue colour arises from the released electrons in the solution when the metal is solvated by ammonia.
Although in this case ammonia is a superior solvent to water, there are some issues which can potentially make the use of ammonia as a solvent awkward for an experiment. Firstly the boiling point of Ammonia is -33.33C  under standard conditions, so the reaction vessel must be cooled to an appropriate temperature (liquid ammonia was first obtained by Faraday in 1823 ). In addition the relative permittivity (the ability of a substance to allow charge to flow through it, measured by the reduction of the Coulombic force when the material is placed between two point charges) of ammonia (16.61) is inferior to that of water (80.100) under standard conditions in effect this means that ammonia will be less effective at dissolving polar compounds.
To investigate the potential using liquefied ammonia as an alternative media to water two syntheses were designed by The School of Chemistry, University of Bristol. Synthesis A makes use of the strong reducing solutions formed when an alkali metal is dissolved in ammonia. The second , synthesis B, takes advantage of the non-aqueous environment to create a complex which would be unattainable in the presence of water.
This synthesis aimed to create 1,2-Bis(diphenylphosphino)ethane, or by its more common name in ligand chemistry dppe, the structure is shown below in figure 1. Dppe is often a bidentate ligand, though can on occasion form unidentate ligands. It is important in the modern world of pharmaceuticals as it is often used, along with other biphosphine ligands, to form complexes which are then used as an asymmetric catalyst to enantiomerically pure drugs.
The key step in this synthesis occurs after the reducing solution has been formed, i.e. after the sodium metal has been added to the liquid ammonia. When the triphenylphosphine is added to this solution it is reduced to Na? [NaPPh2]? . This species is very easily oxidised in air and could be a potential source of error. Further explanation is found in the experimental section below.
Hexa-ammine chromium (III) nitrate (structure shown below in figure 2) cannot be created in aqueous solution as Cr(OH)3 will be formed and the NH3 ligand will not be strong enough to displace it.
The experimental procedure used was provided by the School of Chemistry of Bristol University and was followed with only minor adaptations. The experiment was split into two parts, Synthesis A and Synthesis B. One student performed synthesis A while another performed synthesis B.
Synthesis A was the preparation of bis(diphenylphosphino)ethane, also known as dppe, which structure is shown above in figure 1. The product, dppe, is useful as a bidentate ligand in many inorganic complexes, although there are a minority of complexes where dppe does not act as a chelating ligand.
Synthesis B was the preparation of hexa-amminechromium (III) nitrate, [Cr(NH3)6](NO3)3. The structure of which is shown above in figure 2.
Both syntheses required an atmosphere of ammonia which was achieved by the same method. Firstly a Schlenk tube equipped with a magnetic stirrer bar and septum to a Schlenk line. The Schlenk tube was then evacuated under vacuum conditions and refilled with a Nitrogen atmosphere, hence eliminating possible unwanted side reactions with oxygen, water and other possible reactants found in the standard atmosphere. The Schlenk tube was then placed in an insulated bath containing dry ice and ammonia, resulting in a temperature of - 78C. With the stirrer functional the canula from an ammonia cylinder was inserted through the septum of the Schlenk tube. About 15cm of ammonia was allowed to condense.
Approximately 0.20g of sodium metal was diced into 3 or 4 pieces and added to the Schlenk tube with the septum removed under a strong flow of nitrogen, and then the septum was replaced. Once sufficiently stirred the sodium is oxidised to form a cation and release an electron forming a dark blue coloured solution. The oxidation is shown in equation 1.
Once this solution was obtained 1.14g (4.35 mmol) of triphenylphosphine was added to the reaction vessel, again against a strong flow of Nitrogen. This is the preparation of NaPPh2 as shown in equation 2.
The solution turned a dark brown colour and was stirred for 5 minutes. 0.24g (4.5mmol) ammonium chloride was then added to the Schlenk tube again under a strong flow of Nitrogen, giving a red/orange solution.
0.18cm³ (2.3mmol) of 1,2 dichloroethane was then added to the reaction vessel and was thoroughly stirred for 15 minutes which left an orange solution and small quantity of solid precipitate. This reaction is shown in equation 3 below.
The ammonia was then removed from the Schlenk flask by removing the flask for the insulating bath allowing it to warm and the ammonia to evaporate. This left behind a white solid product, any traces of ammonia were removed by placing the flask under vacuum for about a minute and then refilling the flask with Nitrogen and then turning off the Schlenk line as the product is no longer sensitive to air. The flask was removed for the Schlenk line and the septum removed. About 20cm³ of water was introduced. This formed a suspension which was poured into a separating funnel and 20cm³ of dichloromethane (DCM) was added. The organic layer was separated and the aqueous layer was washed with another 15cm³ of DCM before separating the organic layer again. The collected organic layer was then dried with magnesium sulphate and filtered. The DCM was then evaporated off using a rotary evaporator. The crude yield was calculated (18.55%).
The product was then purified by re-crystallisation using propan-1-ol and the crystals were recovered by Buchner filtration. The melting point of the crystals was then recorded.
Approximately 0.05g of Sodium metal was added to the Schenk flask on the Schlenk line as in the set up described above. This produced a dark blue solution, to which 0.02g of anhydrous Iron (III) chloride (FeCl3) was added. The Iron (III) salt acts as a catalyst for the reaction by rapidly becoming reduced by the free electrons in the Sodium solution. This disperses the blue solution and forms a black solution. 2.13g (7.9mmol) of anhydrous Chromium (III) chloride (CrCl3) was then added to the Schlenk flask under a strong flow of Nitrogen in a few portions to minimise the risk as the reaction is exothermic. The reaction is shown in equation 4 below.
The reaction was then stirred for 10 minutes before the ammonia was evaporated off. This was achieved as in synthesis A by removing the flask from the insulating bath and allowing it to warm to room temperature. Any traces of ammonia wee then removed under vacuum, then the Nitrogen atmosphere was restored and a brown residue was left behind. The flask was then removed from the Schlenk line. *The brown solid was dissolved in 5cm³ dilute HCl with gentle warming. The solution was then filtered. The filtrate was washed with 2cm³ concentrated HNO3 and cooled in ice. This step forms the desired hexa-amminechromium (III) nitrate product. The equation for the reaction is shown in equation 5 below.
The remaining orange product was then Buchner filtered, washing with ice cold water, ethanol then diethyl ether. The product was air dried, an IR spectrum taken and yield recorded (6.67%).
*Note: It was important that the following steps were carried out as fast as possible to try and ensure that the following side reaction did not occur:
This experiment provided 0.17g of the product, which represents an 18.55% yield, the associated error with this yield is ± 2.94% (±1.29x10?³mmol). The error was calculated using a combination of errors formula.
This experiment returned 0.18g of refined product, which represents a yield of 6.67%, the associated error with yield is ± 2.78% (±5.00x10?³mmol). The IR spectrum taken of the product is shown below in figure 3. The analysis for the IR spectrum is found in table 1.
Discussion and Conclusions:
The melting point of the sample produced by this experiment (130.6-131.4°c), was lower than the accepted literature value of 143.5°c , this is an error of 8.71%. A possible explanation for the low yield and low melting point are contamination of the reaction flask with oxygen and moisture from the local atmosphere, allowing unwanted side reactions to take place, which in turn created unwanted products with lower boiling points than the desired product, hence causing the observed boiling point to be lower than expected. Some unwanted products will have been removed in the purification process, decreasing the yield.
Overall, the results from this synthesis were non-conclusive as to whether the reaction was successful in making the desired product. However, it is my opinion that the synthesis was at least partly successful and an unknown proportion of the desired product was created, I base this statement on the partial success of synthesis B, which used a similar method and the calibre of the students carrying out the experiment was similar.
The IR spectrum of the sample would support the conclusion that the reaction appears to have been successful in creating the desired product. However, the yield was a very disappointing 6.67% which would suggest that quite a few unwanted side reactions occurred during the procedure minimising the final yield. Possible explanations of this are most likely going to be down to human error as opposed problems with the experimental procedure. Inaccurate use of the Schlenk line for example may have lead to contamination of the Schlenk flask with oxygen and moisture from the local atmosphere.
The experiment set out to produce a sample of hexa-amminechromium (III) nitrate and in that respect it was a success as supported by the IR spectrum of the final product. However, the yield was extremely disappointing which would suggest either a fault in the experimental plan or, more likely, a shortcoming of the students' technical skill.
I would like to acknowledge Mr James Tharian, my lab partner for this experiment who carried out synthesis A. I would also like to acknowledge Miss Ross Sankey and Mr Michael Huwe, who were the lab demonstrator overseeing the experiment. I must also acknowledge Dr Tom Podesta as the teaching lab manager.
- Handbook of Chemistry and Physics 89th Ed, ISBN: 978-1420066791
- First principles of physics; B. Silliman
- Inorganic Chemistry; Catherine E. Housecroft, Alan G. Sharpe 3rd Ed. 2008 Chp.9 ISBN: 978-0-13-175553-6
- Handbook of chiral chemicals Chp. 9; David J. Ager
- Journal of the Chemical Society, 283: The preparation of di- and tri-tertiary phosphines, 1962 1490-1492; DOI: 10.1039/JR9620001490
- University of Bristol, School of Chemistry, Level 2 Teaching Laboratory 2009/2010 QUICKGUIDE
- University of Massachusetts, Boston; CH371 - Advanced Inorganic Chemistry Laboratory Spring 2009 Exp 5- Liquid Ammonia synthesis of Hexammine chromium(III) nitrate - http://alpha.chem.umb.edu/chemistry/ch371/