Bromine trifluoride, made directly from Br2 and F2, has been known since the beginning of the twentieth century . However, it was rarely used in organic chemistry mainly because of its violent nature when it comes in contact with water or certain organic solvents. Martin was one of the first organic scientists to use BrF3 successfully as a fluorinating agent for octahedral nonmetallic elements (1)  followed by Lemal who used it for the synthesis of hexafluorocyclopentadiene (2) . A few years later, Boguslavskaya and coworkers used it for fluorination of some short chlorinated alkane chains which invariably gave mixtures (e.g., 3 and 4) . Gassen also performed similar chemistry and converted 3-chloropropene into α-fluoroacrylic acid (5)  (Fig. 1).
The ability of bromine trifluoride in replacing halogens and even hydrogens with fluorine was the base for some very popular routes for the synthesis of modern anesthetics. Several papers and patents have appeared describing the syntheses of sevofluorane (6) and desfluorane (7), the latter very stable metabolically and hence practically without any side effects . Some other polyfluorinated ethers were also prepared and showed great potential as anesthetics. Thus, for example, Baker prepared 1,2-bis(fluoromethoxy)-1,1,3,3,3-pentafluoropropane (8) by reacting 2-methoxy-2,2-difluoro-1-(trifluoromethyl)ethyl-fluoromethyl ether (9) with BrF3 (Fig. 2) . This compound also exhibited enhanced metabolic stability with sleep times of approximately six times longer than those for 9.
Additional substitution reactions led to either acyl fluorides (10)  or sulfonyl fluorides including the difficult to make Nafion® sulfonyl fluorides (11) . Another substitution which should be mentioned is the tertiary hydrogen substitution of the adamantine skeleton performed by Hara  (Fig. 3).
In recent years one of the main successes of bromine trifluoride was the preparation of the important CF2 and CF3 groups. Unlike the “brute force” of the above substitutions, one can use the fact that the bromine in BrF3 is a soft acid which easily complexes itself with soft bases such as nitrogen and especially sulfur thus bringing the non-solvated nucleophilic fluoride found in the reagent to the vicinity of the electrophilic carbon (Fig. 4). Such reactions provide an excellent opportunity to turn alkyl bromides into desired corresponding trifluoromethyl alkanes (12)  with or without the positron-emitting 18F isotope (13)  (Fig. 4).
This main idea spanned into a variety of different methods suitable for the synthesis of many other fluorine-containing families. Thus, one of the best ways of making trifluoromethyl ethers (14) from practically any alcohol is to react the latter via the easily prepared xanthates (15) with BrF3 . This could also be achieved by reacting any alcohol with the inexpensive thiophosgene resulting in chloro-difluoro ethers (16). This compound could be treated with tetrabutylammonium fluoride to form the desired 14 (with or without the corresponding 18F)  or with tributyltin hydride to form difluoromethyl ethers 17  (Fig. 5).
The basic idea of replacing sulfur atoms with fluorine by using BrF3 took additional step when it was used for constructing the CF3 group α to a carboxylic acid. A strong base created an anion α to the acidic group followed by the reaction with CS2 and MeI creating a terminal olefinic bis(methylthio) derivative 18. This compound was reacted for less than a minute with the oxidative BrF3 to produce mixtures of more than 85% of methyl 2-bromo-2-[difluoro(methylthio)methyl]alkanoates 19, along with the respective sulfoxides 20, and traces of the sulfones 21. The mixture was treated with HOF•CH3CN (another offspring of F2) at room temperature, transferring all sulfur-containing compounds to the good leaving corresponding sulfones 21 in a few minutes. These were reacted with Bu4NF eliminating both bromine and sulfone groups to give the target α-trifluoromethylalkanoates 22 in overall yields of up to 70% based on the starting esters. Obviously, this method is also suitable for introducing the important isotope 18F into the CF3 group (by reacting 21 with Bu4N18F) for positron-emitting tomography (PET) purposes (Fig. 6) . With secondary carboxylic acids, the reaction is even simpler since no bromine is incorporated in the penultimate step and a yield of overall up to 70% could be achieved (Fig. 7) .
Another approach for constructing the CF3 group is to turn the carboxylic acid itself to the trifluoromethyl group. As with previous examples this approach relies also on the soft basic sulfur atoms which attract the soft Lewis acidic bromine (Fig. 4). The carboxylic acids (both aromatic  and aliphatic ) were readily converted into the corresponding dithioesters (23) and then reacted with BrF3 to produce the aromatic or aliphatic CF3 derivative in 70–75% yield (Fig. 8). It should be pointed out that only aromatic, tertiary, and secondary acids can participate in this reaction, while primary acid produced mainly tars. It is believed that the reason for such behavior is the fact that primary esters are easier to “enolize” (24) and the BrF3 reacts faster with the resulting double bond.
An alternative way to transform alkyl halide to the corresponding trifluoromethyl alkane (25) is via decarboxylative fluorination, a process radical by nature. Thus, alkyl halide was reacted with 1,3-dithiane which in turn was lithiated and brought in contact with ethyl chloroformate . Alternatively direct carboxylation with CO2 could also be carried out forming 26. Reacting such dithiane carboxylic acids with bromine trifluoride resulted in the replacement of the two sulfur atoms by fluorine producing α,α-difluoro acid intermediates (27). When either such intermediate or genuine α,α-difluoro acids served as starting materials, the radical reaction with BrF3 produces the target 25. The reaction does not proceed well with α-monofluoro acids since the monofluoro radical is less stable than the difluoromethylene counterpart, which sustains better radical chain reaction with BrF3. This is also supported by the fact that adding dinitrobenzene, a well-known radical chain interrupter, results in tars only with no detectable trifluoromethyl alkane (Fig. 9) .
As we have seen, the CF3 group could replace halides in alkyl halides, replace COOH group in carboxylic acids, be attached to the α position of aliphatic acids, or be bound to oxygen atoms all by forming variations of appropriate sulfur derivatives and reacting them with BrF3. This strategy can be applied also to form various N-CF3 amide derivatives. The initial approach was to react different imides or amides with tris(methylthio)carbenium tetrafluoroborate (28). Indeed, when this reagent was reacted with the anion of phthalimide (29) or caprolactam (30), the corresponding NC(SMe)3 was obtained and when reacted with bromine trifluoride produced the appropriate N-trifluoromethyl amides (31 and 32) in 60% yield (Fig. 10).
This route, however, is somewhat problematic since 28 is very moisture sensitive and what is more, many amides did not give satisfactory results. An excellent alternative proved to be the use of isothiocyanates (33). These derivatives were reacted with ethanethiol to form ethyl alkyldithiocarbamates (34) in almost quantitative yield following by reaction with carboxylic acids in the presence of N,N′-dicyclohexylcarbodiimide (DCC) and 4-(dimethylamino)pyridine (DMAP). The respective ethyl alkyl(alkanoyl)dithiocarbamate derivatives (35) formed a complex with the bromine atom of the BrF3, and the nearby fluorides efficiently substituted the sulfur atoms forming the desired N-trifluoromethyl compounds (36) in good yields (Fig. 11) .
While the CF3 group is very important in pharmacological industry, the CF2 moiety gained also huge importance recently. The role of the bromine trifluoride in this field was also crucial. Thus, various carbonyls could be readily transformed into their hydrazones, azines, or oxime methyl ethers. As with sulfur, nitrogen atoms are also relatively soft bases thus suitable to complex with the BrF3. The reaction resulted in the appropriate difluoromethylene derivatives (37) in very good yields (Fig. 12) .
As with the chemistry leading to the CF3 group, the sulfur atom(s) found an important role also for the CF2 group construction. Thus, 2-alkyl-1,3-dithiane derivatives, easily made from alkyl bromides and the parent 1,3-dithiane, could be reacted with BrF3 to form the corresponding 1,1-difluoromethylalkanes (RCHF2) (38) or (39) in 60–75% yield. The reaction is suitable for primary and secondary alkyl halides, although with the latter the dithiane has to be prepared in less conventional way. The two sulfur atoms of the dithiane are essential for the reaction (Fig. 13) .
A somewhat similar reaction was developed by Hara who used BrF3-KHF2, an air-stable solid prepared from BrF3 and KHF2. Hara reacted this reagent with benzylic sulfides (40) and in about 15 min obtained α,α-difluoro benzylic carbonyl derivatives (41) in good yields. This reaction was suitable to transform ketones and aldehydes through their dithioacetals (42) to the corresponding aryl difluoromethyl compounds (43). Another important development, unique to this reagent, is the substitution of mono-sulfur starting material such as (phenylthio)glycosides (44) with one fluorine atom to produce fluoroglycosides (45) (Fig. 14) .
Interesting routes were developed for transferring ketones, aldehydes, and alkyl halides to terminal difluoroolefins formally resembling the Wittig reaction (R2C = O to R2C = CF2). The first route is suitable for ketones and aldehydes. Bis(methylthio)methane (46) was reacted with TMSCl and then with the carbonyl derivative to make 2-alkyl-1,1-bis(methylthio)alkene (47). The second route is for alkyl halides and utilize the reaction of tris(methylthio)methane 48 forming 1-alkyl-1,1,1-tris(methylthio)methane derivatives (49). Both compounds, 47 and 49, were brought in contact with BrF3, followed by sulfur oxidation with HOF.CH3CN, to give the difluorosulfonyl derivatives 50 and 51. Consecutive treatment with Zn led to the target difluoroolefins (52 and 53) in overall yields of 60–75% (Fig. 15) .
As we have already seen in Fig. 6, esters could be converted to methyl 2-bromo-2-[difluoro(methylthio)methyl]alkanoates 18 and then oxidized to 21 with the good leaving sulfone part in a few minutes. Compound 21 could be treated with Raney nickel to give various α-alkyl difluoroacrylates (54), some of them potentially promising monomers for future polymers (Fig. 16) .
Reactions of BrF3 could also help in making compounds which were not described before such as α,α-difluoroethers (55) except for one paper with DAST . One example is the conversion of esters to thioesters by using Lawesson’s reagent followed by treatment with the reagent . Another example is the synthesis of aromatic (56)  and aliphatic (57)  difluoromethylenedioxy derivatives (Fig. 17). It should be mentioned that there are several natural products possessing methylenedioxy derivatives and it should be of interest to compare them with their difluoromethylenedioxy counterparts.
One of the problems with BrF3 is the presence of a strong electrophilic bromine which may interfere when reactions are performed on compounds which possess a regular aromatic ring since in addition to what is intended one will get an aromatic bromination as well. The best solution for that problem is to work with the complex pyridine•BrF3 readily made by mixing the two ingredients. Both BrF3 and Py•BrF3 have their advantages and disadvantages. Since the pyridine is complexed to the bromine atom in Py•BrF3, it reduces its ability to act as a strong electrophile, and therefore it does not brominate the aromatic ring but reacts only with the strong soft base such as sulfur or nitrogen present in the target molecule. On the other hand, when bromination is not an issue, the fluoride in the free reagent is a stronger and unhindered nucleophile which may give better results. The following table demonstrates this trend  (Table 1).
Comparing reactions between BrF3 and Py•BrF3 Table 1
As we have seen above, the side product of aromatic bromination can be an issue, but it can be of advantage in cases when the issue is difficult aromatic bromination. Most aromatic bromination processes are carried out in the presence of various Lewis acids, such as AlCl3, serving as catalysts. Their main disadvantage is the requirement for anhydrous conditions and especially for the large amount of catalyst needed, which often results in substantial amounts of effluents discharge, raising serious disposal problems. Furthermore, bromination of aromatic compounds, in particular non-activated ones, can be either impossible or a time-consuming process with relatively low yields. Since the bromine atom in BrF3 is a very strong electrophile, it does not require any Lewis acid, and even the most difficult reactions such as the bromination of dinitrobenzene take place quite efficiently  (Fig. 18).
There are very few reactions with BrF3 which do not involve introduction of fluorine atom into the second reactant. One of them is the aromatic bromination as we have seen above, and the other is transforming primary azides into nitriles in acceptable yields of 30–60%. It is believed that at first a complexation of the electrophilic bromine in BrF3 with the basic nitrogen atom of the azido group takes place and after losing nitrogen and BrF, an unstable RCH2NF2 intermediate is formed, which readily eliminates two molecules of HF to give the more thermodynamically stable nitriles  (Fig. 19). Such elimination is not without precedent as it was reported that even the more stable derivatives such as RCF2NH2 readily eliminate two molecules of HF to produce the corresponding nitriles .
In conclusion BrF3 is a powerful reagent, and while in certain simple cases similar reactions can be replicated with [BrF] (the combination of bromonium source such as NBS and anhydrous HF), it is much more potent and can produce results even when [BrF] is practically inert.
a) Kartashov, A. V.; Chuvatkin, N. N.; Boguslavskaya, L. S. Zhur. Organich. Khim.1988, 24, 2522 (Engl. Trans.: Russ. J. Org. Chem.1989, 2279). b) Kartashov, A. V.; Chuvatkin, N. N.; Boguslavskaya, L. S. ibid1991, 27, 2522 (Engl. Trans.: Russ. J. Org. Chem.1992, 2243). c) Kartashov, A. V.; Chuvatkin, N. N.; Boguslavskaya, L. S. ibid, 1989, 25, 2452.Google Scholar
(a) Rozov, L. A.; Huang, C. G.; Halpern, D. F.; Vernice, G. G. US Patent 5,283,372, 1994. (b) Rozov, L. A.; Huang, C. G.; Halpern, D. F.; Vernice, G. G.; Ramig, K. Tetrahedron Asymm. 1997, 8, 3023. (c) Ramig, K. Synthesis2002, 2627.Google Scholar