Prepare solution of 105g hydrazine dihydrochloride (1 mol) in 250 ml of water (Note #1). Solution is placed into beaker fitted with thermometer for measuring temperature of the liquid, mechanical stirrer and is placed on heating water bath. Solution is heated to 60oC and 84g of dicyandiamide (1 mol) is added in small portions with constant stirring (Note #2). After addition of last portion of dicyandiamide reaction mixture is allowed to sit for 2 hours at 60-80oC (Note #3). Produced transparent solution of guanazole is cooled down to room temperature and 379 ml of 40% nitric acid is added (3 mol, 192 ml commercial 70% HNO3 + 202 ml of water). Reaction mixture is cooled to 0C on ice bath to affect precipitation of guanazole nitrate (Note #4). Precipitate is separated on Buchner funnel and recrystallized from 165 ml of hot water (below 65oC !!!), yielding 135g of guanazole nitrate mixture (Note #5), 72% of theory. Free guanazole can be obtained by neutralization of nitrate salt with alkali, followed by crystallization. White crystalline solid with m.p. 204oC (with decomposition), soluble in water, forms mono and dibasic salts.

Note #1. Reaction only proceeds well in strongly acidic conditions, those are provided by hydrazine dihydrochloride and allow protonation of cyano group in dicyandiamide, as the first stage of reaction’s mechanism. Chloride anions present are known as good promoter of additions to the cyano group and allows high selectivity towards guanazole. Usage of hydrazine monohydrochloride or sulphate dramatically reduces reaction yield.

Note #2. Reaction proceeds with notable evolution of heat, while temperature of the process should not exceed 50-60oC due to formation of notable quantities of byproducts, like 3,5,7-triamino-1,2,4-triazolo[4,3a]-1,3,5 triazine, yield of the later reaches 10-13% at 100oC. Dicyandiamide should be added in portions to control exotherm and to prevent formation of undesirable byproducts.

Note #3. Solution of hydrazine dihydrochloride has notable yellow color, during reaction process it diminishes to the golden color of guanazole salt solution. Reaction greatly speeds up at higher temperatures, at 80-100oC it completes in several minutes, but to reach maximum selectivity and to minimize amount of byproducts it’s better to carry out whole process at 50-60oC in duration of 2 hours.

Note #4. Solubility of guanazole dinitrate strongly depends on temperature, and is illustrated by following numbers (g in 100 ml): 230 (80оC), 158 (70оC), 112 (60оC), 52.4 (40оC), 14.7 (20оC), 5.9 (5оC). Such solubility curve allows separation of guanazole in form of nitrate salt, bypassing normal process requirements (evaporation of solvent in vacuum followed by extraction). Used amount of nitric acid is optimal for separation of maximum amount of salt. Guanazole dinitrate salt is unstable on storage on the air and slowly loses nitric acid.

Note #5. Warning! Attempt of determination of product melting point in thin capillary led to explosion, that occurred before melting (at ~120-130oC) and led to destruction of thermometer. Titration of product with NaOH showed, that product contains 1.4 mol of HNO3 for 1 mole guanazole and is mixture of mono and di nitrates of guanazole. According to reference data, stability of dinitrate is limited by 90oC, leading to decomposition of compound, that caused explosion mentioned above.

Qualitative test for guanazole. Copper (II) sulfate pentahydrate is mixed with same visual amount of sodium acetate and the mixture is dissolved in minimum amount of water, leading to deep blue solution, containing copper (II) acetate; when few drops of the reagent are added to neutral solution of guanazole, light green precipitate is immediately formed. According to reference data its composition is DAT*CuSO4*H2O. Addition of additional quantity of reagent can lead to dissolution of precipitate with formation of dark green solution containing copper-DAT multinuclear complexes. Reaction of hydrazine salt with same reagent leads to markedly different process: at first solution turns blue/violet and either changes color to opaque yellowish with quick liberation of nitrogen and precipitation of metallic copper, or deposits white insoluble precipitate containing copper (I) compound.

To the 300 ml reaction beaker, equipped with dropping funnel and efficient mechanical stirrer (Note #1), placed 55 g of sodium nitrite (0.8 mol), 90 ml water and mixture is heated on water bath at 50oC until solid dissolves. Dropping funnel is filled with solution 10g of 3,5-diamino-1,2,4-triazole (0.101 mol), 125 ml water and 9 ml of concentrated sulfuric acid (0.168 mol, Note #2). Solution of dropping funnel is added drop wise to vigorously stirred solution of sodium nitrite, while maintaining temperature of 50oC at such rate, that whole addition process take around 4 hours. After each hour of addition mixture is allowed to sit for 15 minutes at 50oC before addition is resumed again (Note #3). After addition of each drop yellow foam forms around the drop point, that quickly dissolves in reaction mixture (3 mono and 3,5 dinitrosoamino-1,2,4 triazoles) and some gasses are formed (nitrogen oxides are easily detectable by odor); while at places where released gasses contact with solution of guanazole in addition funnel intense red insoluble solid forms (Note #4). At the beginning of the addition process, reaction mixture is transparent and has lime-yellow color, at the middle of addition it turns into tea color (Danger!!! Note #5). After complete addition of guanazole, stirred reaction mixture is kept at 50oC for 1 hour, when temperature is raised to 60oC and solution is stirred for 1 more hour (Note #6). After aging process, with vigorous stirring 22.7g of 50% sulfuric acid (0.232 mol) is added dropwise during 1-2 hours, on completion mixture is neutralized by baking soda and stirred at boiling water bath for 1 hour (Note #8). Reaction mixture from small amount of colorless solid impurity (polytriazenes) and allowed to evaporate slowly at room temperature (Note #9). Crude mixture of salts (consisted mostly of Na2SO4*10H2O, NaDNT and some NaNO3) is extracted by 150 ml of the acetone (150 ml), filtered from insoluble material, 15 ml of water is added and extract is allowed to evaporate slowly at room temperature (Note #10). Sodium salt of 3,5-dinitro-1,2,4-triazole shows some isomorphism with sodium sulfate, that leads to occlusion of product in crystals of sodium sulphate, so for more complete isolation of product acetone extraction is repeated 2-3 more times, with though grinding of solids before each extraction (Note #11). After 3-x extraction by acetone 10.2g (64% of theory, ref [4] gives 66% yield) of sodium salt of 3,5-dinitro-1,2,4-triazole is collected in form of red-orange crystalline material (Note #12). Separation of free DNT from the sodium salt was described in refs [3,4].

Note #1. According to data from ref [4] yield and purity of dinitrotriazole are improved when mechanical stirring was used instead of magnetic one, and also on slower addition of guanazole solution.

Note #2. Selection of sulfuric acid (instead of HCl or HNO3, often used for such reactions) is based on several reasons. First one is its inability to enter the redox processes (direct substitution of diazonium group require electron transfer from anion), leading to higher purity of product than in case of HCl (in later case product is contaminated by chloro-triazoles). Second is that it’s diazonium salts are more stable than extremely sensitive and unstable diazonium nitrates, and precipitation of relatively low soluble guanazole nitrate salts is the feeding line is prevented.

Note #3. Employed synthetic procedure is different to normal Sandmeyer diazotizations preformed in presence of copper salt catalysts at low temperature and is instead the chain radical process driven by Srn1 mechanism [1,2,3]. Possibility of direct substitution of diazonium group by this mechanism is based on the fact that redox potentials of nitrite ion is close to the classical Cu(2+)/Cu(+) redox pair, that provide electron transfer in Sandmeyer and Gatterman’s reactions. Formation of diazonium salt is fast process already at 0oC, while process of replacement of diazonium group with nitrite ion is much more slow and proceeds readily only at temperatures around 50oC. So, the method described above, allows to combine diazotization and replacement processes at the same time, allowing to mitigate accumulation of highly dangerous and unstable diazonium intermediates, that are present at the reaction mixture. Slow addition rate of guanazole solution and 15-minute cooldowns serve the same purpose.

Note #4. Nature of this product is not completely clear, it’s insoluble in water and is probably a triazene formed by azo-coupling of diazonium salt (formed by reaction of escaping nitrous gases with guanazole solution) with pristine molecules of guanazole. Authors of work [5] describe several coupling reactions of guanazole, and from visuals and properties product in question corresponds to compound IV – 3,5-bis{[3-(5-amino-1,2,4-triazolyl)]triazenyl}-1,2,4-triazole. Under the microscope compound appears as small orange-red crystals, having high affinity for water. Taking to account visual similarity it’s still unclear, why exactly this products form or it is something else entirely. Whatever is the case, formation of this product is undesirable as it will lead to contamination of target compound, while diazotization out of boundary of main reaction solution leads to danger of crystallization and explosion of formed diazonium salts, so effort should be taken to minimize contact between escaping nitrous gases with guanazole solution in addition funnel.

Note #5. DANGER! Approximately at the middle of addition period, strong micro-detonation occurred, accompanied by visible white flash, this was without a doubt the decomposition of tiny amount of unstable diazonium compound. It is well known that stability of diazonium salts greatly depends on anion and composition of aromatic substrate: electron-donating groups in the ring and voluminous non-oxidizing anions like tetrafluoroborate or sulfate provide stabilizing action, while electron-withdrawing groups in the core greatly destabilize diazonium salts. Guanazole contains two amino-groups; those can be diazotized or be replaced by nitro group in the same time leading to formation of extremely unstable and dangerous diazonium salts rivaling, instability of the 5-diazotetrazole. Inspection of damage shown explosion occurred on the surface of mechanical stirrer and was probably caused by contact of drops of guanazole solution with nitrous gasses, followed by drying the stream of air at the stirrer rod’s surface. Event again places emphasis on importance of safety measures and equipment, that should be employed whole time and especially during guanazole addition process.

Note #6. Purpose of the aging process is to maximize substitution completion and decomposition of unstable diazonium intermediates. After complete addition solution was transparent, had color of dark tea (with some reddish tinge), and contained only small trace of crème-white precipitate (polytriazene).

Note #7. Addition of sulfuric acid induces sizeable form formation, caused by formation of nitrogen oxides, originating from decomposition of excess of sodium nitrite. After addition of most acid release of gasses and foaming are stopped, mixture was fully transparent, had lime-yellow color and show pH <= 1.

Note #8. Reaction equations show, that synthetic procedure should spent 3*0.1*0.5 = 0.15 mol of [H+], while 0.4 was added in total, so neutralization should require 0.25 mol of hydrocarbonate (21g). In reality neutralization took 19.4g, and caused color change to tea color, formation of trace of insoluble precipitate, separated by filtration through the cloth, pH = 7-8.

Note #9. Literature data on DNT salts [6] suggest high sensitivity of anhydrous salts. Slow evaporation at room temperature insures formation of safe, hydrated product. Evaporation to dryness on 4 plates at room temperature took around 3 days. Examination of product under microscope shown large amount of rod habit crystals (sodium sulfate octahydrate) and yellow DNT salt in form of plates and needles. After complete drying some small amount of impurity is seen in from of star like agglomerates of orange-red color.

Note #10. Intentional addition of water is aimed on separation of product in the form of dyhydrate. Fast evaporation of sample of acetone solution leads to formation of pike shaped golden crystals of product observed under the microscope. According to reference [6] anhydrous salt is colorless, while hydrate is yellow-orange. Material obtained on evaporation of solvent with added bit of water is orange-yellow microcrystals.

Note #11. Part of allotropic forms of hydrates of Na2SO4 and the anhydrous salt can form monoclinic crystals showing some isomorphism with monoclinic crystals of NaDNT and it’s dehydrate, that leads to easy occlusion of DNT salt by hydrated sodium sulfate crystals. Yield after one extraction is only 35% from theory, so extraction is conducted multiple times. Major part of solid residue are Na2SO4*10H2O (dec.temp. 35oC), Na2SO4*7H2O (dec.temp. 25oC) with some amount of anhydrous salt. Sodium sulfate have anomalous solubility curve below 40oC (during cooling from 25oC to 0oC solubility is reduced 10 times and is around 5g of salt in 100 ml at 0oC) that can be used to separate most sodium sulfate before the next extraction. Solution of this salt have notable habit for form supersaturated solutions, that suddenly crystalize as a whole on lowering the temperature and have extremely fluffy structure and are hard to filter. So it’s advantageous to perform cooling from 40oC to 0oC very slowly, adding some seed crystal, that leads to formation of huge (several centimeter long) crystals of sodium sulfates that occlude minimal amount of mother liquor and DNT salt. For more complete separation of the later, solid residue from first extraction is ground to fine powder and extracted by new portion of acetone, undissolved solid is dissolved in minimal amount of boiling water (around 150 ml) cooled to 40oC, seed crystal is added and solution is slowly cooled to 0C, precipitated crystals of Na2SO4*10H2O are separated by filtering and mother liquor is evaporated at room temperature, grounded and extracted by acetone the third time. Yield after first extraction is around 30%, second extraction gives 25% more, and third one gives more 10%. Total yield reaches 10.2g of NaDNT (64% of theory).

Note #12. Crystallization from anhydrous acetone and from acetone with addition of water leads to product containing some minor amount of (like 1-2%) of ruby-red crystals clearly visible under microscope. This impurity is highly soluble, since it forms on the outer edges of NaDNT crystals. Nature of this impurity is unknown.

References:
[1] Aromatic substitution by the SRN1 mechanism; Joseph F. Bunnett; Accounts of Chemical Research 1978 11 (11), 413-420; DOI: 10.1021/ar50131a003
[2] Radical nucleophilic substitution mechanism in the reactions of arenediazonium cations with nitrite ion; P.R.Singh, Ramesh Kumar, R.K.Khanna; Tetrahedron Letters Volume 23, Issue 49, 1982, Pages 5191-5194; DOI: 10.1016/S0040-4039(00)85794-9
[3] Heterocyclic nitro compounds. I. Synthesis of nitro derivatives of 1,2,4-triazole, 1,3,4-thiadiazole, tetrazole, 1,3,4-oxadiazole, and pyrazole by the noncatalytic replacement of the diazo group by the nitro group ; L.I. Bagal, M.S. Pevzner, A.N. Frolov & N.I. Sheludyakova; Chemistry of Heterocyclic Compounds volume 6, pages 240–244 (1970) ; DOI: 10.1007/BF00475005
[4] Preparation and characterization of 3,5-dinitro-1H-1,2,4-triazole ; Haiges, R. and Bélanger-Chabot, G. and Kaplan, S. M. and Christe, K. O.; Dalton Trans. V44, I.16, P 7586-7594 (2015); DOI: 10.1039/C5DT00888C
[5] The Diazotization and Autocoupling of Guanazole; M. Hauser; The Journal of Organic Chemistry 1964 29 (11), 3449-3450; DOI: 10.1021/jo01034a532
[6] Synthesis and structural characterization of 3,5-dinitro-1,2,4-triazolates; R. Haiges, G. Bélanger-Chabot, S. M. Kaplan, and K. O. Christe; Dalton Trans. 2015 Feb 21;44(7):2978-88; DOI: 10.1039/c4dt03583f.

Experiment №1. 25.2g of dicyandiamide and 40.1g of ammonium chloride were thoroughly grounded and mixed together and placed into iron can from the juice with top cut open (volume was approximately 240 ml). Can was placed into stand holder and was heated by the open flame of gas oven. No visual change was observed during first 10 minutes, some white fumes of sublimed ammonium chloride were notable. Temperature of mixture was observed during stirring, by means of the thermocouple imbedded into glass stirring rod. After temperature reached 160oC, layer of liquid melt was observed at the places of contact of mixture with iron can surface (where heat transfer is better), after this big exotherm sets up with almost immanent rise of temperature to 230C, after this whole mixture was liquefied into mobile liquid. After exothermic reaction started heating was discontinued and mixture was left to cool down, causing solidification of reaction mass. Solid product was exceptionally hard and could be hardly broken by strikes of hammer. Solid mass was dissolved in 150 ml of boiling water, leading to almost complete solution with almost no insoluble material (almost complete absence of ameline). Separation of solid impurities by filtering led to clear transparent solution with slight yellowish tinge. Next, solution of copper (II) sulfate in concentrated ammonia was added until solution achieves blue coloration (excess of copper tetramine sulfate). Additions of first few drops of solution led to formation of reddish-crème coloration, that was almost immediately discharged and changed into dark blue. Filtering afforded 0.5g of brownish-red crystals of unknown composition. Absence of copious amount of precipitate with copper (II) tetramino sulphate, shown almost complete absence of biguanide in reaction mixture.

Target compound was not separated, while repeating above procedure, given in reference [1]. Details of method from reference are however uncertain due to inability to replicate heating conditions of Bunsen burner flame, depending from gas flow and heat transfer conditions. Investigation of literature data shown, that strong heating (even in small duration) what took place in the experiment (around 180oC) leads to guanidine with only trace quantity of biguanide, while stronger and longer heating leads to ameline as major reaction product. No attempt was made to separate the guanidine product, however it was clear a major product of above experiment and should have formed in nearly quantitative yield. It was decided to repeat the synthesis by using careful heating on liquid petroleum paraffin bath.

Experiment №2. 25.2g of dicyandiamide and 40.1g of ammonium chloride were thoroughly grounded and mixed together and placed into iron can from the juice with top cut open (volume was approximately 240 ml). Can was placed into stand holder and was heated by molten petroleum paraffin heating bath. When heating bath temperature of 170oC was reached, heating was discontinues and resumed only when temperature dropped below 150oC. In those conditions petroleum paraffin is somewhat volatile and produces some white smoke and is readily detectable by odor, along with some mild odor coming from reaction mass. Heating in bath at 150-170oC for 30-40 minutes had not produced noticeable changes in the reaction mixture, no liquid melt areas were seen. After approximately 1 hour of heating some liquid areas began to form, after 1.5 hours whole mixture was melted, forming viscous liquid similar to honey in consistency. After complete melting mixture was heated to 150-170oC for 15 more minutes, during this time viscosity dropped and mixture became easier to stir. Heating was discontinued and liquid melt was poured onto paper and was grounded before full solidification. Resulting mass was dissolved in 150 ml of boiling water, most of solid dissolved, but some small amount of ameline residue was also present (weighted around 3g). Filtrate was completely colorless solution, that on addition of ammonical copper (II) tetramino sulfate immediately produced huge quantity of pink precipitate, solution remained completely colorless after addition of each portion, before the excess of the reagent was added, leading to deep blue coloration of solution. Full precipitation took about a half of theoretical amount of ammonical copper (II) tetramine sulfate. Filtering afforded copper biguanide sulfate in fair yield as rose-red solid, what is washed with water and dissolved in 35 ml of 10% sulfuric acid (the temperature should not exceed 90°). The resulting solution is then cooled by immersion in an ice bath. The crude crystals which form are separated and dissolved in 25 ml. of boiling water and the solution is cooled in ice. Again the crystals which form are dissolved in 25 ml. of boiling water and the solution is cooled in ice. The resulting colorless crystals of biguanide sulfate 2-hydrate are filtered, washed first with 10 ml. of cold water and then with 10 ml. of absolute ethanol, and dried at 110° for about 15 hours. The yield of anhydrous biguanide sulfate is 10.1 g. (16.9%); melting point 232oC.

Properties: Pure anhydrous biguanide sulfate is a white amorphous solid that is not very soluble in cold water but is soluble in hot water, from which it crystallizes in colorless glistening rhombic crystals. An aqueous solution of biguanide sulfate is acidic to litmus. The compound does not dissolve in absolute ethanol to any appreciable extent and is insoluble in common organic solvents. Reaction of compound with theoretical amount of NaOH in dry methanol allows separation of pure biguanide free base, with 90% yield, m.p. 134oC (decomposition at 138oC).