# Reduction of Organic compounds

I've encountered in chemistry quite a lot of reducing agents. However one question strikes me quite often and any research done on the same hasn't been fruitful. This is in context to organic chemistry. Now for example, in reductive ozonolysis, we often use ozone in combination with zinc and water. Zinc's presence ensures reductive results and not oxidative ones. Now, some sources(not books) which I have seen, say that some sulphur compounds would do the same like $$\ce{Me2S}$$. Also, in case of 1,3- dinitrobenzene, $$\ce{Sn + HCl}$$ converts both the $$\ce{-NO2}$$ groups to $$\ce{-NH2}$$ groups. Now if we use $$\ce{NH4HS}$$ or some other nitrogen and sulphur containing ammonia like compounds, we get reduction again, but only for one group out of the two. So my question is, how do we :

1. Decide that there is going to be reduction
2. Know the extent of reduction (partial/full)

Can some insight into the inorganic properties of the elements provide an answer like d-orbital availability, lone pair, electronegativity, size/charge ratio? I've tried reasoning but haven't gotten anywhere. Or is it that memorization of the reagents is the only way to know it's function.

Note: This question is only in context to some less commonly used reagents, so as to recognize it's function if we ever see one in problems present in question books.

• You already underestimated it, a single reagent can give many products in proportions depending on various conditions. In chemistry, experiment tells what does what and when and conducting it in seemingly identical conditions can give surprising effects. – Mithoron Mar 8 '20 at 21:38

I will try to explain the difficulty of this problem by using a simpler reaction, the reduction of nitrobenzene, which is already difficult enough. The reduction of dinitrobenzene should produce more different possibilities.

Nitrobenzene can be reduced into at least five different compounds. It depends on the choice of the reducing agent. I am afraid nobody can justify the particularly mechanisms used by a given reducing agent to get to each final result.

With ($$\ce{Sn + HCl}$$) or with ($$\ce{Fe + HCl}$$), it gives aniline, $$\ce{C6H5NH2}$$.

With sodium methoxide ($$\ce{NaOCH3}$$), it gives azoxybenzene, $$\ce{C6H5N(O)NC6H5}$$

With zinc, it can produce three different products:

1. With zinc + $$\ce{NH4Cl}$$ at elevated temperature, it gives phenylhydroxylamine, $$\ce{C6H5NHOH}$$.

2. With zinc and an alcoholic solution of $$\ce{NaOH}$$, it gives azobenzene $$\ce{C6H5N=NC6H5}$$.

3. With zinc and an aqueous solution of $$\ce{NaOH}$$ at $$\pu{60 ^\circ C}$$, it give hydrazobenzene $$\ce{C6H5NH-NHC6H5}$$.

• Yes. I understand that a little change in the reagents makes a difference. But can we not check the strength of reducing agents in organic chemistry? The real aim is to determine: 1. Whether it is a reducing agent or not and 2. If yes, then is it strong or weak? i.e., will it be able to reduce all possible positions or only one like in case of NH4HS. – Tesla's Coil Mar 9 '20 at 6:16
• You check these looking at the literature and by experiment. Solvent, temperature and additional reagents make so much difference that there is no systematic rule. If you want to be able to predict everything from equations study physics. – Waylander Mar 9 '20 at 9:54
• @PrarabdhShukla Then you check standard redox potential, but it doesn't tell what and how it can actually reduce. – Mithoron Mar 9 '20 at 16:17
• @Waylander Nice answer. Thanks for the clarification. And yeah, physics isn't all predictable. Classical mechanics is, but quantum isn't . And hence we have chemistry, completely unpredictable (to some extent it is) because we don't where the electron is! So the door for studying physics is closed. But I like both! – Tesla's Coil Mar 9 '20 at 18:10
• Possibly if you went into a lot of data collection, you could try using principal component analysis or PLS, but by the time you got the reduction potential data, solvent properties and other variables collated , you could have done the experiments quicker. All this would be substrate dependent too. – Beerhunter Oct 9 '20 at 20:05