Benzotriazole and derivatives

As benzotriazole is now a widely available and relatively cheap chemical used in the domestic environment as a corrosion inhibitor in circulating water heating systems I thought it would be interesting to see what chemicals could be prepared from it. There are three possible modes of attach that I can see, the first is the imino hydrogen on the triazole ring, the second is direct electrophilic substitution of the benzene ring and the third and potentially most useful is the cleavage of the benzene ring to give a substituted 1,2,3 triazole derivative (see The Chemistry of Heterocycles by T. Eicher and S. Hauptmann).

The imino hydrogen tautomerically distributed between the 1 and 2 positions so alkalating agents tend to give a mixture of 1 and 2 substituted product though 1 generally predominates. Other reactions tend to attack the 1 position preferentially. The imino group can be halogenated to give 1-halo derivatives which in common with other haloimine (ie N-bromosuccinamide and TCC) acts as halogenating agent.
The second reaction seems fairly limited as the benzene ring is not particular reactive and so requires fairly aggressive conditions to bring about substitution, however it can be directly nitrated in the 4 position and concentrated hydrochloric acid an oxidizing agent such as nitric acid give 4,5,6,7 tetrachlorobenzotriazole.

The most interesting reaction, however, is the oxidation with potassium permanganate which cleaves the benzene ring and generates 1,2,3-triazole-4,5-dicarboxylic acid. This compound is potentially the starting material for many other compounds, it is likely to lose a carboxylic acid group fairly readily to give the 4-monocarboxylic acid. Both the mono- and di- carboxylic acid should esterify by the Fischer process in methanol and the esters can be converted to the amide and thence to the nitrile. Though by analogy with pyrazine-2,3-dicarbonylamide the dehydration to a vicinal dinitrile requires aggressive conditions (thionyl chloride) and yields are poor. There may also be a possibility of using Hoffman’s rearrangement of the amide to access the mono-amine though again by analogy with the pyrazine equivalent it will probably not work with diamide (Paul E. Spoerri & A. Erikson JACS 1938 p400).

The 4,5-dicyano-1,2,3-triazole is an energetic material and acts as a monobasic acid, it is usually prepared from the very inaccessible diaminomaleonitrile (aka hydrogen cyanide tetramer C4N4H4) and nitrous acid.

So here’s the results of my first attempts at oxidizing benzotriazole to the triazole dicarboxylic acid.

My 1,2,3-Triazole-4,5-dicarboxylic acid Preparation

The reaction is assumed to proceed in accordance with the following equation:

6KMnO4 + C6H5N3 -> C4HN3O4K2 + 2K2CO3 + 2H2O + 6MnO2

160g + 20g = 39g + 46.4g + 6g + 87.6g

And:

C4HN3O4K2 + 2K2CO3 + 6HCl -> C4H3N3O4 + 6KCl + 2H2O + 2CO2

39g + 46.2g + 36.6g = 26.3g + 74.8g

20g of technical grade benzotriazole were dispersed into 400ml of water at 50°C and then 150g of finely powdered potassium permanganate added in very small portions of about 1g at a time to the mechanically stirred mixture in a 2L beaker.

The initial reaction generates considerable heat and the purple colour of the permanganate is rapidly replaced by a brown precipitate of manganese oxide. Initially there is considerable foaming and the stirrer must be run at a fairly high speed to help break up the foam. It also helps to rinse down the sides of the beaker every so often with a wash bottle. If the temperature exceeds 75°C, which it did several times, add a little crushed ice and suspend addition of the permanganate until the reaction mixture has cooled a little. The addition took about 2 hours but after the first 30 minutes the foaming ceased.

Once addition of the permanganate was complete the dark brown slurry was stirred until almost cold, no residual permanganate was left so the slurry was filtered at the pump and rinsed with about 50ml of cold water. The manganese dioxide cake was preserved and dried for chlorine production.

The light brown filtrate, about 470ml, was boiled with 2g of decolourizing carbon and filtered again while still hot to give a clear straw coloured solution. This was evaporated down on a water bath to about 80ml by which time it had acquired a dark brown colour so was treated with a further 1g of decolourizing carbon and filtered hot. On cooling the light brown solution deposited long, almost colourles blades. The solution was warmed until the crystals dissolved and transferred to a 250ml beaker and acidified with 30% hydrochloric acid until the solution was distinctly acid to Congo Red paper. There was much foaming at first as the potassium carbonate was decomposed and then the solution because much paler and deposited a flock-like precipitate. The small amount of flock like precipitate was filtered off and the pale straw coloured solution reduced to about 100ml on the water bath and cooled overnight. Masses of minute, pale straw coloured prisms formed which were recovered at the pump and sucked as dry as possible. They are rather soluble so no attempt was made to wash then on the filter; the yield of crude dicarboxylic acid was 8.63g of pale straw coloured crystals when dry. These crystals melt when heat on a spatula and burn away leaving a trace of white soluble residue (KCl).

The filtrate was evaporated do to roughly half its original volume (to about 50ml) and allowed to cool. A heavy almost white granular crystalline precipitate formed that looked very different from the first crop of crystals. The crystals were practically infusible on a spatula and decrepitate badly. Chemical test reveal that it is mostly KCl.

The first crop of crystals has not yet been re-crystallized.

I intend to try this oxidation again but using a lower temperature and a slight excess of potassium permanganate to see if I can improve the yield. I also intend to see if I can esterify the acid.

Has anyone else experimented with this material

The reaction mixture after addition of 100g KMnO4 and some ice

After filtration and evaporating down to 80ml prior to 2nd Norit treatment

Quote: Originally posted by Boffis

Well I can't attach it because it over 2MB so if anyone want to look at it contact me by U2U

The story continues:

Having had further time in the laboratory I have now recrystallised the 8.6g of crude straw coloured triazole dicarboxylic acid. I found that the material can be recrystallized with good recovery from water using 5ml per g. 8.63g were recrystallized from 40ml of water with the addition of 0.1g of decolourising carbon and filtering through a preheated buchner funnel followed by cooling, final in the fridge overnight. 7.48g of pure white product were recovered after drying.



Straw coloured triazole dicarboxylic acid 1st recrystallisation

Pure white acid after 2nd recrystallization

The liquor left from the recrystallization was not discarded but was kept and with be worked up with the solution that resulted from boiling the Mn oxide cake with 300ml of fresh water.

Nicodem referred to a patent that desribed the oxidation of various heterocyclics with hydrogen peroxide, nitric acid and ferric nitrate as a catalyst. I downloaded the patent and decided to give it a try. I used the technique described in the patent almost to the letter except that I used ten times the quantity and added the nitric acid in 2 goes uncase the reaction turned violent; it didn't. I will post the details of the experiment later but basically after neutralizing most of the nitric acid a mass of yellow crystals formed. When recrystallized from water the filter cake consisted of a mixture of two products. An almost white fibrous material which appears to consist entirely of unreacted benzotriazole and a second compound as blocky yellow crystals in the ratio of about 2:1. I have been unable to seperate them by recrystallization but when recrystallized from water in the presence of a little EDTA the crystals of both are practically coourless so the colour is probably due to iron. I will continue my investigations of this reaction and the product later this month.



Left; solution after oxidation for 18hrs & Right; crystallization after neutralizing
Left; the raw filter cake & Right; recrystallized material filter cake, the 2 phases clearly visible

Quote: Originally posted by Boffis

Picric acid + PCl5 => Picryl chloride Picryl chloride + NH3 => 2,4,6 trinitroaniline

Quote: Originally posted by Boffis

4,6 dinitro 1,2 phenylene diamine (by analogy with formation of picramic acid) 2,4 DN oPD + NaNO2 + HCl => 4,6 dinitrobenzotriazole

Quote: Originally posted by Boffis

I decided to try the hydrogen peroxide oxidation of benzotriazole again but this time using less and stronger nitric acid. I will also re-run the permanganate oxidation.

Quote: Originally posted by Boffis

Alternatively it may be possible to obtain this material via picric acid. Picric acid + PCl5 => Picryl chloride Picryl chloride + NH3 => 2,4,6 trinitroaniline 2,4,6 TNA + NaS => 4,6 dinitro 1,2 phenylene diamine (by analogy with formation of picramic acid) 2,4 DN oPD + NaNO2 + HCl => 4,6 dinitrobenzotriazole

Quote: Originally posted by Boffis

One last thought that I am working on that someone out there may be able to comment on (Nicodem!). I have acquired some 2-nitro-1,4-diaminobenzene dihydrochloride (used in hair dyes), its rather old but looks reasonably good (brownish crystals). If I reduce the nitro group to an amine to give me 1,2,4 triaminobenzene do you think I will be able to react it with one molar equivalence of sodium nitrite+HCl to give 5-aminobenzotriazole? In other word are all three amino group equally reactive towards nitrous acid or will the ortho pair react preferentially? There is also an organic synth preparation of 4-nitro-1,2-diaminobenzene from 2,4 dinitroaniline which looks like a possible route to 5-nitrobenzotriazole.

Hi Boffis,
I think that in 1,2,4-triaminobenzene, the 3 NH2 are equivalent towards diazotation but the formation of the triazole ring is very fast.
In typical aromatic amine diazotation with excess aromatic amine, one usually get a side reaction of azo-coupling...that's why NaNO2 and HCl are used in slight excess otherwise aniline diazotizes and react in para position of another aniline molecule.
Ar-NH2 + HONO --> Ar-N=N-OH + H2O
Ar-N=N-OH + Ar-NH2 --> Ar-N=N-Ar-NH2 + H2O
In the case of o-diaminobenzene, as you have observed, no coupling occurs between o-diaminobenzene and o-amino-benzodiazonium because the intramolecular NH2 reacts very fast with the viccinal diazonium and cyclise into benzotriazole before it has a chance to react extramolecularly.

You will thus have diazotation in the 3 positions 1, 2 and 4.

1°)Positions 1 and 2 are equivalent because the derivated benzotriazoles are resonant.
-C-N=N-NH-C - <==> -C-NH-N=N-C -
So actually 4-amino-benzotriazole is also 5-amino-triazole.
In the case you work stoechiometrically, you would end up with 2/3 of 4-amino-benzotriazole.

2°)Position 4 may also be diazotised, if you work stoechiometrically, the resulting 3,4-diamino-benzodiazonium will react with another identical molecule in position 6 (only free para position to the NH2 in position 3, position 4 is oposed to the diazonium).
The resulting molecule can be polymeric in nature:

-Ar(NH2)2(-N=N)-)n

but it can also be dimeric

(H2N)2Ar(-N=N-)2Ar(NH2)2

Indeed the two ortho diazo bridges can interact to form a tetraaminodibenzotetraazapentalene related to tetranitrodibenzotetraazapentalene (TACOT). Further diazotation would then add two triazole rings at the extremities because each end of the molecule displays ortho-diamino groups.



or trimeric

((H2N)2Ar(*)(-N=N-)Ar(NH2)2(-N=N-)Ar(NH2)2(-N=N-)(*))
(the * binds together to cyclise the molecule)

In a triangular fashion with aromatic rings on each tops (with two ortho NH2 groups on each aromatic ring pointing where the triangle top points to) and with diazo bridges as the sides of the triangle to join two aromatic rings 2 by 2.
Here again further diazotization will induce triazole ring extension to the ends of the triangle.

3°)If exces HNO2 is used then you will end up with 4-diazonium-benzotriazole

[Edited on 3-8-2014 by PHILOU Zrealone]

@ Philou Zrealone

Your post raises some fascinating possibilities that I was not aware of or had not thought about; I have now had time to think about this a good deal and done a fair bit of research too. I am currently far from home so no experiments are possible at present but when I return home I will try the reduction of 2-nitro-phenylenediamine and then diazotizing it with a single molar equivalence of nitrous acid or perhaps a more (x2) and see what we get! My thinking on the resultant mixture is that simple diazo linkages can be reduced back to amines, so the ortho diazo linked benzotriazole compound should be reducible to 5,6 diamino-benzotriazole with something like sodium dithionite so allowing the recovery of some potentially useful by-product, the problem may be separating the product efficiently. However, see my comments below about reaction rates of competing coupling reactions.

Theory is one thing but some experimental data is required here so I will give this reaction a go when I have time at home. By the way the resulting aminobenzotriazole would be the 5 isomer;



I have also discovered an interesting synthesis (2) of a fused quinoline-triazole-arsonic acid from the 1,2,4 triamine using toluenesulphonyl substitution to protect one of the o-amines. This reaction was carried out in a single step, the arsonic acid entering the 4 position! This suggests that the diazotization of the free o-amine group and its cyclotization to a triazole occurs rapidly with the cleaving of the tosyl group faster than it can condense with the arsenite ions leaving only the 4-amino position to undergo Bart's arsonic acid synthesis. This suggests that it may be possible to protect both amino groups in my 2-nitro-1,4-diaminobenzene with tosyl chloride, then reduce the nitro group with say H2S or ferrous ions (avoiding strongly alkaline and acid conditions which might hydrolysis the tosylamine groups) to 1,4-ditosyl-1,2,4-triaminobenzene and then diazotize the new amino group to give the 5-tosylaminobenzotriazole. Any comment on this idea? Since I have the materials I'll give this a shot too when I have time.

An alternative has just occurred to me; Griess (6) described the coupling of diazotized sulphanilic acid with an o-diamine to give a triazole. Since cyclotization reactions of this type are usually fast and the products stable it may be possible to react diazotised sulphanilic acid with 1,2,4 triaminobenzene in HCl solution to get mainly the 5-aminotriazole because the competing reaction, coupling to the 4-amino site to give an initial aminoazo compound which will then probably rearrange to an o-amino-azo compound, is likely to be slow in an acid solution (think of the formation of aminoazobenzene and its rearrangement to p-aminophenylazobenzene). The initial coupling requires weakly acid conditions and rearrangement occurs in more acid condition because the initial compound dissociates into the diazonium compound and amine again and the second o-coupling reaction is much slower so only competes in an acid solution. It may be possible to inhibit the 4 and 5 coupling reactions completely with sufficient acid. Furthermore it may be than the diazo-cyclotization of the 1,2 diamine occurs so readily that it limits the scope for diazotization of the 4-position amine to a minor by-product. I certainly feel that there is scope in this reaction to get a reasonable yield of a triazole.

While thinking about the preparation of other benzotriazoles I came across a reference to the synthesis of 1-Hydroxybenzotriazole via the condensation of hydrazine with 2-nitrochlorobenzene, Sauron had discussed this reaction back in 2007 as a possible route to 2-nitrophenylhydrazine and thence to 1-hydroxybenzotriazole but it looks like the forcing condition required cause the hydrazine moiety to react with the nitro group give the hydroxy triazole in a single step. This reaction is referred to in Eicher & Hauptmann (p207) but unfortunately the reference at the end of the paragraph is to the preparation and use of simple hydroxy triazole derivatives (1) and I can't track down a reference for this reaction. However, this reaction raises a whole series of possibilities e.g. could 2,4-dinitrochlorobenzene (of which I have plenty) be subject to a similar condensation to give 5(6)-nitro-1-hydroxybenzotriazole and 2-nitro-1,4-dichlorobenzene (available from p-dichlorobenzene + sulphuric acid + alkali nitrate) to give 5 or 6-chloro-1-hydroxybenzotriazole. I have found a reference to this reaction with 2,4-dinitrochloro-naphthalene with phenylhydrazine to give 2-phenyl-5-nitronaphtho(1,2)triazole-3-oxide so there is some merit in this idea (3) and other references to similar reactions of active o-halogens with hydrazines have also been found; some examples results in a simple triazole while other in a mixture of a triazole plus a 1,2-bi-substituted hydrazine. Nitro groups on the benzene ring can be reduced to amines and these active this ring to further substitution, these reactions were extensively covered by Fries and Zincke ((7) many references).

Has anyone come across the reaction of 2-nitro-aromatic amines with hydrazine? The significance of such a reaction is its attack via the nitro group and therefore overcome the problem of which amino group in 2-nitro-1,4-diaminobenzene will react, I vaguely remember seeing such a scheme once but I might be mistaken. An internet search revealed only reduction to amine but in the presence of large amounts of Mg (the hydrazine seems superfluous to me), I have found a reference to the reduction of 2-nitrobenzenediazonium compounds to the N-hydroxybenzotriazole with hydrazine; the use of sulphite gives the hydrazine so conditions must be important (4).

@ Anders:- I have found a reference to the preparation of 4,6-dinitrobenzotriazole from picramide via 3,5-dinitro-1,2-diaminobenzene as I speculated about above and it does work in HCl (5).

All of this, however, is moving away from the spirit of my original idea which was to explore the compounds that could be prepared from commercially available benzotriazole and methylbenzotriazole through either substitution or degradation of the benzene ring. I haven't yet had time to follow up the esterification of the acid or its partial decarboxylation followed by esterification to the monoacid ester. From these esters I intended to attempt the preparation of the amides followed by either Hoffman's degradation to the amine or dehydration the nitrile.

I have also discovered another oxidative degradation of the benzene ring in benzotriazole via intensive chlorination and hydrolysis/oxidation to give 4,5 di-substituted triazoles with chlorinated or undersaturated substituents, there are many reference to this by Fries and Zincke in J.L. Annalen Chem and Berichte, it will take some time to check these out and see which one are the interesting ones (7).

One curious thing is that I have not found a single reference to simple dimethyl or diethyl esters of the 4,5 dicarboxylic acid and only very few to any ester at all; then mostly to N-substituted phenyltriazole carboxylic acids. If anyone has access to a chemistry specific search I would appreciate any references you can find on their synthesis and properties.

@Philou: Do you have a reference to TACOT?

(1) Reference in; The Chemistry of Heterocycles 2nd ed - T. Eicher, S. Hauptmann:- M. Begtrup, P. Vedso, Acta Chem Scand, 1996, v50 p549; This refers to the synthesis of N-Hydoxy-1,2,3-triazoles and 4 or 5 substituted analogues.
(2) Slater, R. H.; J Chem Soc; 1932, pp2196-7
(3) Mangini, A.; Atti accad. nazl. Lincei, Classe sci. fis. mat a nat. 25, 326-332 (1937) I have not been able to find this reference
(4) temporarily lost; to follow
(5) Nietzki, R. and Hagenbach, H.; Berichte. 30, pp544-545 (1897)
(6) Griess, P. Berichte; 1882; v15; p2199
(7) see refs in Benson F. R. & Savell, W. L.; Chemistry of Vicinal Triazoles; Chemical Review, 1950 p1-68

Quote: Originally posted by Boffis

Several members have asked me where the tenzotriazole cames from; well it is available once again on Ebay, in the UK at least, at a very reasonable price in quantities from 200g to 2kg. At 17 GBP for 500g its pretty cheap.

Quote: Originally posted by Pumukli

Interesting what you do with these triazoles and unveiling a possible otc source would be a nice teaser too. :-)

Interesting stuff you said there regarding the azo-amino rearrangment. The interaction between different substitutions on the ring and the behaviour of the corresponding nitro/nitroso/nitr(so)amine/(di)azo(xy) groups under different conditions is amazingly complex. the more you read about it, the more complicated it seems to become. My guess would be that the p-amino group of 1,2,4-triaminobenzene is likely to lead to a lot of side reactions, since only in the case of an o-diamine the intramolecular rearangement after diazotization to the triazole would be favored. How stable is triazole to NaHS reduction and basic conditions, guess there is a reason you want to go via the triamino?

Regarding the 4-substituted o-phenylenediamines, would benzimidazolone be a good starting point? n-methyl benzimidazolone can be found OTC. The two para positions relative to the aminogroups would be most activated. Hell, it might even nitrate all the way to the tetra nitro and produce an tetranitrotriazole.

[Edited on 16-12-2015 by nitro-genes]

Nitration of Benzotriazole

I have finally gotten around to experimenting with the nitration of benzotriazole, its only taken 5 years. This is still work in progress and I need to prepare the 5-nitro isomer by another method, as described above, to check the identity of the two main products. I then plan to reduce it to the amine and investigate the interesting array of derivatives described in the references at the end and others not included yet. I also intend to investigate the possibility of further nitration to 5,7-dinitrobenzotriazole which can also be prepared from picric acid via picramide (2,4,6-trinitroaniline) and 3,5-(4,6)-dinitro-o-phenylenediamine 4) as discussed in earlier posts. I then intend to post the final write-up in the Pre-publication section and this version will contain translations of the more important German references.

Nitration of Benzotriazole
Boffis January 2022

According to the work of Fries et al1) benzotriazole is nitrated exclusive in the 4 position e.g. the position adjacent to the fused triazole ring. The 5 isomer is easily prepared by indirect means from 2,4-dinitroaniline via the selective hydrosulphide reduction to 4-nitro-1,2-phenylene-diamine2) and in better detail3).


The method described below was developed from the method of Fries et al (p229) and worked admirably but TLC of the product revealed that it is a mixture of 3 compounds. One of them, an orange-coloured material occurs in only trace amounts but the other two are significant components, the purification process suggests a ratio of about 3:1. The more abundant component forms tiny, blocky, very pale-yellow crystals and is presumed to be the 4-nitrobenzotriazole while the minor component forms slightly dark yellow fibrous aggregates and fits Zincke’s description of the 5-nitrobenzotriazole but could be in part at least a dinitro compound. The two compounds are easily separated due to their large difference in solubility, particularly in acetone and butanone.

Experimental
Preparation scale nitration of commercial benzotriazole

72.12g of crude commercial benzotriazole granules were added fairly quickly to 205ml of technical grade concentrated sulphuric acid with gentle stirring. The benzotriazole dissolves quickly and the temperature rises to about 60°C. The light brown solution was cooled to about 5°C and then placed in an iced water bath on a stirrer plate (note 1). 44ml of concentrated (69%) nitric acid (approximately 10% excess) were added slowly from a dropping funnel at such a rate that the temperature did not rise above 30°C.


Figure 1; crude benzotriazole (left) and benzotriazole dissolving in sulphuric acid (right).

If the temperature starts to rise too quickly place the beaker in a freezer for 20-30 minutes until the temperature has dropped to about 0°C and then continue the addition of nitric acid. When the addition was complete the beaker was allowed to stand at room temperature for 20 minutes and then warmed to 55-60°C slowly over 30 minutes to complete the reaction.

The clear amber coloured solution was allowed to cool to 15°C and poured over 100g of ice and the beaker rinsed with about 25ml of cold water. The hot, pale yellow, suspension was allowed to cool to room temperature (about 8°C) and 200ml of 50% sodium hydroxide added slowly over about 5 minutes (note 2). The suspension become exceedingly hot and was allowed to cool again to room temperature overnight (to about 5°C).


Figure 2; fully dissolved benzotriazole solution in H2SO4 (left) and after addition of the nitric acid.


Figure 3; the reaction mixture after heating for 30 minutes, ready to drowned and the drowned reaction mixture.

The suspension was filter using a large (12.5cm) Buchner funnel, washed with a little water and then pressed down and sucked as dry as possible, any cracks in the cake healed with a spatula. The pale straw-yellow cake was dispersed into 350ml of cold water and then stirred for 15 minutes until a thin cream-like suspension had formed. The suspension was filtered again and sucked as dry as possible and then air dried at 35-40°C to constant weight on two 200mm watch-glasses, this took nearly 2 days to achieve. The yield was 80.81g or 81.3% of theory (note 3). A TLC revealed that the raw product is a mixture of three compounds. A faint orange line visible without UV is only a trace impurity but the other two are both significant components.



Figure 4; filtering off the nitration product, note the clear filtrate (left) and the still-moist cake of crude product.

The crude nitrobenzotriazole was purified by the following method: 38.5g of crude product were boiled in a 500ml conical flask with 195ml of acetone (5ml per g) briefly and then the suspension allowed to cool to room temperature. The pale pinkish yellow crude 4-nitrobenzotriazole was filtered off washed with a little (25ml) acetone and dried to give 25.347g. The filtrate, about 220ml, was distilled to recover 110ml of acetone and cooled to room temperature overnight. The gritty, pale-yellow crystals of fairly pure 4-nitrobenzotriazole were filtered off and dried to give a further 3.318g of 4-nitrobenzotriazole. The remaining filtrate was distilled to give about 50ml of orange liquid that was poured into a glass bowl and allowed to cool and evaporate. When evaporation was complete and the smell of acetone gone a pale orange granular material filled the bottom of the bowl while around the edge was a brighter pinkish orange efflorescent crust. The granular material was carefully removed leaving the efflorescent rim adhering to the edge of the bowl. The granular material is believed to be mainly 5-nitrobenzotriazole and weighed 8.436g. The efflorescent rim was carefully removed and weighed 0.601g appears to be a mixture of the 5-nitro isomer and a delicately fibrous orange compound. The missing 0.8g is the result of mechanical and handling losses that are difficult to avoid with hot acetone solutions because of their low viscosity and surface tension and the need to work quickly.

The crude 4-nitrobenzotriazole can be recrystallised from boiling acetone or butanone (20ml per gram) to give very pale-yellow blocky crystals in 55% yield without working up the filtrate. Other solvents invariably give a woolly fibrous product that is much less convenient although the single-crop yield may be better e.g. 18ml per gram of 95% ethanol gives a single-crop yield of about 76% but of fine, matted fibres. Work is still in progress on the two minor products but methanol (6ml per gram) appears to be the best solvent for recrystallising the granular 5-nitro isomer, the yield is about 90% but it still appears to be a mixture of scaly rosettes (the major product) and tiny fibrous pompoms that resemble the 4-nitro compound when it is recrystallised from ethanol.

Note 1; With small scale experiments using 20g of benzotriazole ice-water is adequate but for this larger scale preparation it would be better to use an ice-salt bath and probably cool the starting solution to -5 or below first.
Note 2; nitrobenzotriazole is fairly soluble even in dilute sulphuric acid, the amount of sodium hydroxide neutralises most of the sulphuric acid to sodium hydrogen sulphate greatly reducing the solubility of the base. Further neutralisation to neutral sodium sulphate is best avoided since large amounts of it crystallise out complicating the work-up and reducing yield through mechanical losses as I have discovered in earlier experiments.
Note 3; Attempts to recover more material from the original sodium hydrogen sulphate filtrate proved worthless, less than 0.3g were recovered by further neutralisation and decanting from the sodium sulphate crystals through a filter. The mixture of nitrobenzotriazole and sodium sulphate hydrate being washed with a little water and the residue dried.

1) K. Fries, H. Güterbock and H. Kühn; Annalen der Chem.; Investigations into the series of benzotriazole and N-methyl-benzotriazole; p213-240 (1934)
2) A. W. Hofmann; The action of nitrous acid on nitrophenylenediamine; Annalen der Chem; v115, p249, (1860)
3) T. Zincke; On Chloroketones and quinones of heterocyclic compounds and their derivatives; Annalen der Chem., v311, is3, pp276-329 (1900)
4) Norton & Elliot; Dinitro-1,2-diaminobenzene; Berichte v11, p327 (1878)

@ AvBaeyer; many thanks for your helpful comments. I have now played around and found that chloroform-methanol seems to work well and I have settled for a 5:1 ratio. Interestingly this mixture has resolved the original nitration product and, more clearly, the final residue from recrystallisation into 5 spots!

I am very please with my final 4-nitroproduct which gives only a single spot. I am now working on recovering the other product which I presume to be mainly the 5 isomer.

I forgot to mention that the TLC I used in the original prep were developed with pure ethyl acetate but this causes streaking so the upper streak can only be resolved into 2 bands. The lower streak from the origin to about 0.4 is the yellow impurity. With the chloroform methanol solvent it resolves into a single fairly sharp band at about .15 so I'm pleased .