[ravipeace]

Hi,  I have to find the heat of reaction in dehydration of Basic Ammonium Aluminium Sulphate in which 2 water molecules are released at about 450 K.

Reaction is like this.  A.8H2O = A.6H2O+2H2O

where A.8H2O is the Basic Ammonium Aluminium Sulphate.

One way is, I must know the enthalpies of formation of each compound at given temperature. I dont even know the standard enthalpies of formation except for H2O.

Or, there must be other way like some literature where a similar exercise has been performed.

Any help will be appreciated.

Regards,

Ravi Gupta

Research Scholar
#Mechanical Engineering.

[Plontaj]

estimated H of formation NH4*Al(SO4)2*8H2O in 25oC equals about -1400 kcal per mole
estimated S of formation NH4*Al(SO4)2*8H2O in 25oC equals about 135 cal per mole per K

estimated H of formation NH4*Al(SO4)2*6H2O in 25oC equals about -995 kcal per mole
estimated H of formation NH4*Al(SO4)2*6H2O in 25oC equals about 115 cal per mole per K

I found only NH4*Al(SO4)2*12H2O in database but estimated values and values from database are not so different (error for H less than 2%).

[Enthalpy]

I don't see any good method except measured values. Either find the accurate heat of formation of the hexahydrate and octahydrydrate, or find the right heat of dehydration, or measure it by yourself.

For a reaction where little heat is expected, a difference between two estimated values is far too inaccurate. Here the claimed -5858kJ/mol octahydrate, -4163kJ/mol hexahydrate, 2*-242kJ/mol for water would leave 1211kJ/mol, unbelievable: that would be the heat of formation of five water molecules from hydrogen and oxygen.

So much inaccurate in fact, that you can abandon your estimation method and seek a better one. A heat of hydration is just a few times the heat of fusion of water, not its heat of formation.

Relative uncertainties aren't meaningful for heats of formation because we use their differences.

[Plontaj]

Due to I was in the hurry I made mistake during writing it down. I'm sorry, I attached a screen shoot with the "right" values.

[Enthalpy]

May we know how this estimate was made?

Because it suggests exactly the same -71.6kcal/mol change in enthalpy of formation for each added H₂O, be it from 6 to 8 or from 8 to 12, which would be sort of a surprise.

I also feel that the resulting hydration enthalpy, -13.8kcal per mol of H₂O, would be a lot. As a comparison, the trihydration of Al₂O₃ releases only 17.8kJ per mol of H₂O.

[Plontaj]

The calculation has been made in HSC. Based on their database (containing over 28k chemical compounds) has been made an algorithm to predict the enthalpy and enthropy.

http://www.outotec.com/en/Products--services/HSC-Chemistry/Calculation-modules/H-S-Cp-estimates/

[Enthalpy]

That's what I had expected. The same heat of hydration for successive additions of water molecules just stinks after software.

Software prediction does NOT work well enough for the enthalpy of formation, and far less so when the sought reaction involves little heat like a hydration does. The "mean error" of HSC's estimate is 10kcal/mol, letting expect 100kJ/mol error often, gasp. An even worse example for software prediction is the melting point, which is completely off.

Welcome in nomal life, where Nature does not obey programmers.

[Plontaj]

No one said, that estimated values are better than experimental. In case when we have no information about physicochemical properties always rough values are better than nothing.

[Enthalpy]

Then the proper way is to measure. The least bad is to take experimental values from similar compounds.
Whether estimated values are better than nothing when they're off possibly by a factor-of-4? I use to prefer no information at all.

[Plontaj]

Sometimes to preparing the concept of technology is not necessary to know all "true" values that will be used in that particular technology. Especially at initial evaluation of economic aspects of technology, then unvalidated data are useful as well; sometimes are demanded.

[Enthalpy]

When evaluating a feasibility too, correct figures are vital. A factor-of-4 error isn't acceptable.

My opinion is that something should automatically tag
>>>>> BEWARE SOFTWARE, DON'T TRUST <<<<<
every time a value is not measured but comes from:
Estimation software
In silico experiment
Additive rules
First principles
...and the like

It's a real plague on the Web, and even now in science publications, that people put wrong values and don't even tell they don't come from the real world.