1

Teadustöö kirjeldus 2014 Agnes Kütt Sihtasutus Eesti Rahvuskultuuri Fond Tehtud teadustöö tulemused võib välja tuua järgmiste punktidena:

1) 2013 aastal avaldasime artikli (ChemPlusChem, 2013, 78, 932–936), milles kirjeldati diasooniumsoola C6(CF3)5N2

+ BF4– tegemist. Antud sool sünteesiti

vastavast aniliinist C6(CF3)5NH2. Aniliini süntees ei ole triviaalne, selle tegemine käib läbi 4 etapi ning aine lõplik saagis lähteainetest võiks olla suurem. Kuna antud diasooniumsoola vastu tunti huvi eesmärgiga kasutada seda pindade funktsionaliseerimiseks, siis seda ainet oli vaja suuremas koguses. Sellest tekkis idee proovida sünteesida sama aniliini teisiti: 1) iodoneerida nitrobenseen (saaduseks C6I5NO2), 2) viia läbi trifluorometüleerimisreaktsioon (saaduseks C6(CF3)5NO2) ning seejärel lihtne katalüütiline hüdrogeenimine (saaduseks C6(CF3)5NH2). Kuna trifluorometüleerimisreaktsioon ei andnud ootuspäraselt vastavat nitrobenseeni vaid hoopis üllatuslikult fenooli (Joonis 1), siis antud teemat uuriti lähemalt ning sellest sai ka artikkel (RSC Adv., 2014, 4, 41895–41901). Leiti, et kui C6(CF3)5NO2 ja C6(CF3)5NO segule (tahkena türkiissinise värvusega) lisada Me4NOH lahust, värvub lahus punakaks (mida põhjustab ilmselt tekkinud NO2 ja NO) ning tekib üks C6(CF3)5 rühma sisaldav aine. MS mõõtmistest tuli välja, et see aine on fenolaat C6(CF3)5O–. Kui trifluorometüleerimisreaktsioon viidi läbi pentaiodonitrobenseeniga, et saada C6(CF3)5NO2, siis ei saadud mitte vastavat nitrobenseeni vaid jällegi puhas fenool C6(CF3)5OH. Sarnaselt reaktsiooniga tetrametüülhüdroksiidiga asendab nitrorühma solvendist pärit hapniku aatom. Kui fenooli saamine varemalt meie uurimisgrupis Tartu Ülikoolis väga hästi ei õnnestunud (enamik fenoolist valmistati Bremeni Ülikoolis), siis nüüd on suurepärane meetod suurema koguse vastava fenooli sünteesimiseks. Samas kahjuks ei ole võimalik niiviisi saada nitrobenseeni C6(CF3)5NO2, mis võiks olla suurepäraseks lähteaineks aniliini, fenooli ja teiste ainete saamiseks.

Joonis 1. Sünteesietapid nitrobenseenist pentakis(trifluorometüül)fenoolini.

2

2) Aniliini C6(CF3)5NH2 tehes saadi väikese koguse kõrvalproduktina amiin [C6(CF3)5]2NH. See on huvitav amiin selles mõttes, et lämmastikul olev prooton on amiini kohta äärmiselt happeline. Tekkiva happe aniooni negatiivne laeng on delokaliseeritud benseenituumadesse ning lisaks veel anioonne tsenter steeriliselt varjestatud suurte CF3 rühmasid täis benseenituumade poolt. Seetõttu võib happe aniooni (amiidi) pidada ka nõrgalt koordineeruvaks aniooniks. Antud amiin pakkuski rohkem huvi ning töötati välja selle aine süntees vastavast aniliinist ja klorobenseenist (Joonis 2). Aniliini tuleb sünteesi käigus deprotoneerida piisavalt tugeva alusega, kuid huvitav on seejuures see, et alus ei tohi alus olla ioonne alus (BuLi, NaH, BuOK jne). Alus peab olema tugev orgaaniline alus! Ioonne alus reageerib alus klorobenseeniga andes pentakis(trifluoromeüül)benseeni ja reageerimata aniliini. Sama meetodit rakendades oli võimalik sünteesida ka analoogne metaan [C6(CF3)5]2CH2. Amiid [C6(CF3)5]2N– reageerib solvendis klooriga ning tekib klooramiin [C6(CF3)5]2NCl. Kui antud aine panna kokku raua puruga atsetoonis ja ultraheliga aktiveerida siis tekib intensiivse rohelise värvusega lahus (Joonis 3). Kuna on teada, et aminüülradikaalid on rohelised, siis tekkis kahtlus, et niiviisi on võimalik valmistada stabiilne aminüülradikaal [C6(CF3)5]2N*. Antud roheline lahus omas EPR spektrit ehk seal oli paardumata elektron kuid EPR spektri nihke põhjal see ei saanud olla vaba aminüülradikaal. Ilmselt on tegu raua ja aminüülradikaali kompleksiga. Isegi kui roheline värvus lahusest kadus ja lahus jäi kollaseks, siis sama tugev EPR signaal oli siiski olemas. Tekkiv kompleks tundub lahuses olevat stabiilne. Töö eesmärgiks oleks saada antud ainest kristallid, et röntgenstruktuuranalüüsi abil saada täpselt teada, mis aine on tekkinud. Kristallid on umbes pool aastat rahulikult kasvanud – mida kauem kristallid kasvavad, seda paremaid mõõtmistulemusi on oodata.

Joonis 2. Sünteesitee amiinini ja erinevad meetodid amiini 1 modifitseerimiseks. Täpne 1•Fe struktuur ei ole teada.

3

Joonis 3. Väga väikese kontsentratsiooniga [C6(CF3)5]2NCl lahus värvub atsetoonis raua puru juuresolekul ultraheliga aktiveerides roheliseks. 3) Kuna kirjanduses oli palju viiteid selle kohta, et pindade funktsionaliseerimine ei

ole võimalik selliste benseendiasooniumsooladega, mis sisaldavad orto-rühmasid benseenituumas, siis valmistati lisaks punktis 1 kirjeldatud diasooniumsoolale ka 3,4,5-(CF3)3-C6H2-N2

+ BF4– sool. Kuna antud diasoonium sool ei ole niivõrd

elektronpositiivne kui varem kirjeldatud soola süntees, siis selle valmistamine oli üsnagi lihtne vastavast aniliinist. Aniliin omakorda valmistati katalüütilise hüdrogeenimise teel trifluorometüleeritud nitrobenseenist (Joonis 4). Sünteesi kõige keerukam osa oli see, et iodeerides nitrobenseeni tekkisid di-, tri- ja tetra- asendatud iodobenseenid, mida lahutada ei olnud võimalik. Seda segu trifluorometüleeriti ning tekkivad bis-, tris- ja tetrakis- asendatud trifluorometüül nitrobenseende segust sai kuidagi kristallimise teel lahutada tris-derivaadi, kuid seda üsna madala saagisega. Siiski piisav kogus ainet saadi kätte ning see saadeti edasi pinnauurimise grupile Iirimaal. Lõpikke tulemusi, kuidas pindade funktsionaliseerimine on edenenud, ei ole veel teada. Mitmed meetodid, mida katsetati, kahjuks ei andnud positiivseid tulemusi.

Joonis 4. Sünteesietapid 3,4,5-(CF3)3-benseendiasooniumsoolani.

4) Järeldoktorantuuris alustasin püüdega valmistada nitreeniumkatiooni, mis on

stabiliseeritud asendatud karboraan aniooniga. Idee oli panna klooramiin reageerima karboraan-stabiliseeritud silüüliumkatiooniga, tekkima peaks nitreeniumkatioon. Cl11- ja Br6-asendatud karboraanid (valemitega vastavalt CB11Cl11

– ja CB11H6Br6–) on tõestatud väga head vastasioonid mitmetele muidu

ebastabiilsetele katioonidele. Algul valmistati klooramiin asendamata difenüülamiinist (C6H5)2NCl. Vastav nitreeniumkatioon (C6H5)2N+ oli aga liiga

4

ebastabiilne ning saadi hoopis ammooniumkatioon (C6H5)2NH2+. Edasi prooviti

valmistada (2,4,6-Me3-C6H2)2NCl, see oli stabiilne toatemperatuuril umbes pool tundi ning selle aja jooksul oli võimalik see reageerima panna ka silüülium karboraaniga. Tekkisid ka oranžikad kristallid (kirjanduse põhjal peaks nitreeniumkatioon olema värviline). Nende edasine analüüs aga muutus võimatuks, kuna kristall muutus väljaspool kuivboksi kohe vedelaks värvusetuks tilgaks. Suure tõenäosusega ei õnnestunud seetõttu ka röntgenstruktuuranalüüs. Järgmisena sünteesiti kirjanduse põhjal uudne aine: (C6Me5)2NH, millest valmistatud klooramiin oli aga stabiilne vaid mõned sekundid. On ilme, et mida elektronegatiivsem on lämmastik difenüülamiinis seda stabiilsem on klooramiin. Kuna klooramiini valmistamine ei õnnestunud, viidi läbi läbi katse, kus võeti amiinilt (C6Me5)2NH ära vesiniku anioon kasutades Ph3C+ karboraani soola (Joonis 5). Tekkis uut tüüpi aine, millest saadi ka kristallid, kuid kristallid ei olnud piisavalt kvaliteetsed, et röntgenstruktuuranalüüsiga lokaliseerida vesinikke (Joonis 6). Saadi küll info, et kristallid sisaldasid (C6Me5)2N osakesi, kuid kas tegu oli nitreenium katiooniga või ammoonium katiooniga, seda ei ole veel teada. Ka teised analüüsimeetodid (NMR, IR) ei anna selgust, kuna tekkiv aine sisaldab ka solvendimolekule muid reaktsiooniprodukte. Töö antud teemal veel on veel käimas.

Joonis 5. Sünteesitee ostetud lähteainetest nitreeniunkatioonini. {Br6} sümboliseerib karboraananiooni CB11H6Br6

–.

5

Joonis 6. Röntgenstruktuuranalüüsi tulemus, millel on näha Br6-karboraan, (C6Me5)2N osakesed ja solvendi 1,2-diklorobenseeni molekul.

5) Dr. Toomas Rodima sünteesib Tartu Ülikoolis äärmiselt tugevaid neutraalseid aluseid – fosforüliide. Nende lähteaineks on erineva struktuuriga fosfiinid. Fosforüliidide aluselisust mõõdetakse gaasifaasis ning tetrahüdrofuraanis. Mitmete fosfiinide aluselisuse mõõtsin ära aga atsetonitriilis. Mõningad fosfiinid on üsnagi tugevad alused jäädes ligikaudu püridiini aluselisuse kanti. Mõõtmistulemused on ära toodud tabelis 1. Tulemused on veel avaldamata.

6

Tabel 1. Mõõdetud alused ja tulemused. Alus Referents alus pKa(ref) ∆pKa pKa(alus) Lõplik

pKa(alus)

Mes3P Pyridine 12.53 -0.37 12.90 12.88 2-Me-Pyridine 13.32 0.40 12.92 4-MeO-Aniline 11.86 -0.95 12.81 [2,4,6-(MeO)3-C6H2]Ph2P 11.76 -1.12 12.88

[2,4,6-(MeO)3-C6H2]Ph2P Pyridine 12.53 0.77 11.76 11.76 Mes3P 12.88 1.12 11.76 4-MeO-Aniline 11.86 0.11 11.75

(3-MeO-C6H4)3P 2,5-Cl2-Aniline 6.21 -1.03 7.24 7.24 2-Cl-Pyridine 6.79 -0.46 7.25 2,6-(MeO)2-Pyridine 7.64 0.41 7.23

[3,5-(MeO)2-C6H3]3P 2,5-Cl2-Aniline 6.21 -0.97 7.18 7.19 2-Cl-Pyridine 6.79 -0.41 7.20 2,6-(MeO)2-Pyridine 7.64 0.44 7.20 NaphtPh2P 7.31 0.14 7.17

[2,4,6-(MeO)3-C6H2]2PhP 2-NO2-4-CF3-C6H3-P1(pyrr) 16.54 0.65 15.89 15.86 2,3-(NH2)2-Pyridine 15.24 -0.53 15.77 2,4,6-Me3-Pyridine 14.98 -0.92 15.90 [2,6-(MeO)2-C6H3]3P 17.23 1.37 15.86

[2,4,6-(MeO)3-C6H2]3P 4-CF3-C6H4-P1(pyrr) 20.16 0.55 19.61 19.61 2-Cl-C6H4-P1(pyrr) 19.07 -0.55 19.62 2-Cl-C6H4-P1(dma) 20.17 0.57 19.60

CyHexPh2P 2-Me-Aniline 10.5 0.27 10.23 10.23 2-MeO-Pyridine 9.93 -0.30 10.23 1-Naphtylamine 9.77 -0.46 10.23

NaphtPh2P 2,6-(MeO)2-Pyridine 7.64 0.39 7.25 7.28 2-Cl-Pyridine 6.79 -0.48 7.27 [3,5-(MeO)2-C6H3]3P 7.19 -0.14 7.33

[2,6-(MeO)2-C6H3]3P 2-NO2-4-CF3-C6H3-P1(pyrr) 16.57 -0.68 17.25 17.23 [2,4,6-(MeO)3-C6H2]2PhP 15.86 -1.37 17.23 2-NO2-5-Cl-C6H3-P1(pyrr) 17.27 0.05 17.22

RSC Advances

COMMUNICATION

Publ

ishe

d on

19

Aug

ust 2

014.

Dow

nloa

ded

by U

nive

rsity

of

Tar

tu o

n 22

/09/

2014

10:

14:2

6.

View Article OnlineView Journal | View Issue

Pentakis(trifluoro

Institute of Chemistry, University of Tartu, R

agnes.kutt@ut.ee

† In memory of Alexander A. Kolomeitsev

‡ Electronic supplementary informaticoordinates and structures of compoundsof 1 and 1a, 19F NMR spectra of the reaESI-HRMS, and HPLC-ESI-MS spectra ofliquid nitrogen trap. See DOI: 10.1039/c4r

Cite this: RSC Adv., 2014, 4, 41895

Received 14th May 2014Accepted 19th August 2014

DOI: 10.1039/c4ra04540h

www.rsc.org/advances

This journal is © The Royal Society of C

methyl)phenol fromNitrobenzene†‡

Agnes Kutt* and Ilmar A. Koppel

In the present work, a two step method for obtaining pentakis(tri-

fluoromethyl)phenol from commercially available starting materials is

presented. It has been prepared previously using different methods

but all these methods suffer from low yields and/or time-consuming

purification processes. Until now, a large amount of pure pentakis-

(trifluoromethyl)phenol has been difficult to prepare. Sufficiently pure

pentakis(trifluoromethyl)phenol is now obtained from pentaiodoni-

trobenzene in 58% yield and after only one time-consuming subli-

mation process. The phenol forms during a trifluoromethylation

reaction from pentaiodonitrobenzene; iodides are substituted by tri-

fluoromethyl groups, and the nitro group is substituted by an oxygen

atom in the oxygen-nucleophilic environment. A small amount of

pentakis(trifluoromethyl)phenol is available from the authors on

request.

Phenols are very common reagents in chemistry. Tri-uoromethylated and polyuorinated phenols are startingmaterials for ethers,1 esters,2 substituted olens,3 sulfonates,4–6

sulfamates,4 etc. Alkoxy aluminates,7,8 a class of weakly coordi-nating anions,9 are specically prepared from phenols. Themore acidic the phenol is, the weaker is the coordinating abilityof anions. Other types of coordination compounds,10 cata-lysts,11,12 etc., for which phenols are important reagents, arepublished as well.

Pentakis(triuoromethyl)phenol 1 (Scheme 1) is a verystrong acid. Its acid strength (pKa(MeCN)¼ 10.46)13 is similar tothe acidity of picric acid (pKa(MeCN) ¼ 11.00).14,15 Pentakis(tri-uoromethyl)phenolate 1a has well delocalized negative charge,and the most nucleophilic site of the anion is hidden between

avila 14A, Tartu, 50411, Estonia. E-mail:

.

on (ESI) available: Computational2, 3, and nitrobenzene; UV-vis spectractions and products; IR-ATR, GC-MS,products; photograph of blue N2O3 ina04540h

hemistry 2014

two protective ortho-CF3 groups reducing its nucleophilicity.Phenol 1 is a solid (m.p. 88–90 �C) yet volatile substance with arecognizable smell of vanillin that makes the initial detectionsimple. In the case of increased availability of phenol 1, thiscompound is envisaged to nd several useful applications.

Until now, three different methods have been published forobtaining phenol 1. None of these methods have very highyields and some of them include rather inconvenient and time-consuming purication procedures. These methods are:13,16

(1) Triuoromethylation of hexaiodobenzene with pregen-erated CuCF3. Phenol 1 is obtained as a by-product with anapproximate yield of 20%. However, on using N-methyl-2-pyrrolidone (NMP) instead of 1,3-dimethyl-2-imidazolidinone(DMI) only traces of 1 are obtained. An increase in the forma-tion of 1 in DMI is related to the enhanced oxygen-nucleophilicity of DMI compared to NMP. DMI has an approxi-mately 10 times higher price than that of NMP, which makes italmost impossible to use it in large scale triuoromethylationreactions, where it is difficult to recover used solvents.

(2) Hydrolysis of pentakis(triuoromethyl)chlorobenzene 5,yield 38%. The experiments were carried out by another oper-ator, and only 17% yield was obtained aer a difficult puri-cation process that contained several sublimation procedures,which are very time-consuming.17 The yield probably dependson the scale of the reaction; it is faster and simpler to sublimesmaller amounts of the product.

(3) Reaction of pentakis(triuoromethyl)benzenediazoniumcation C6(CF3)5N2

+ with water. The yield of pure 1 from

Scheme 1 Numeration of compounds.

RSC Adv., 2014, 4, 41895–41901 | 41895

Scheme 2 Suggested reaction pathway from nitrobenzene 2 to

RSC Advances Communication

Publ

ishe

d on

19

Aug

ust 2

014.

Dow

nloa

ded

by U

nive

rsity

of

Tar

tu o

n 22

/09/

2014

10:

14:2

6.

View Article Online

pentakis(triuoromethyl)aniline 4 is 39%, similar to theprevious method and aer several sublimation procedures. Thepreparation of aniline 4 from chlorobenzene 5 has a maximumyield of 79%. Therefore, the yield of 1 from pentakis(tri-uoromethyl)chlorobenzene 5 is approximately 31% throughtwo steps. Chlorobenzene 5 is prepared from pentaiodo-chlorobenzene. The entire process takes 4 steps fromcommercially available compounds.

In addition, phenol 1 was also prepared from pentachloro-phenol and pentabromophenol by substituting chlorine orbromine atoms, respectively, by CF3 groups via a tri-uoromethylation method, which is also used in the presentwork.13 CuCF3 did not react with these substances, and it wasnot possible to obtain phenol 1 using these starting materials. Ithas been reported that bromo-18,19 or even chloro-substituted20,21

aromatics can be triuoromethylated using a CF3� source and

different catalytically active metals but these methods areknown to be un suitable for substituting halides in electron-richarenes.19,21

From the reaction of pentakis(triuoromethyl)benzene-diazonium salt with dry NaNO2,16 the mixture of pentakis(tri-uoromethyl)nitrobenzene 2 and -nitrosobenzene 3 wasobtained. It was very complicated to separate these twocompounds because of their very similar volatility and solu-bility. When Me4NOH (20% in MeOH) was added into the NMRtube to the mixture of compounds 2 and 3 dissolved in CDCl3,the solution turned red in color, and the product, according tothe 19F NMR spectrum, was a single almost pure pentakis(tri-uoromethyl)phenyl derivative (See Fig. 1 and spectrum S15 inESI‡). The mass-spectrometric analysis of the product thatcrystallized from CDCl3 revealed that the obtained compoundwas pure phenolate 1a (Spectrum S19‡).

Fig. 1 19F NMR spectra that shows how the mixture of pentakis(trifluoCDCl3 (blue spectrum, a) are converted mainly into a single pentakis(trMe4NOH (25% in MeOH). ESI-HRMS measurements revealed that the co

41896 | RSC Adv., 2014, 4, 41895–41901

Formation of pentakis(triuoromethyl)phenolate 1a frompentakis(triuoromethyl)nitrobenzene 2 appears to be anucleophilic substitution (Scheme 2). Nitrobenzene 2 is anaromatic compound that is fully substituted by highly electro-negative groups – ve CF3 groups and one NO2 group make thebenzene ring extremely electron decient. Thus, partial positivecharge on the aromatic carbon atoms is high. Ipso-carbon of 2has an especially positive partial charge and is eager to bindnucleophiles. In case there are OH� species, strong nucleo-philes, in the solution of nitrobenzene 2 or nitrosobenzene 3,then these attack the ipso-carbon, i.e. the carbon atom with thebest leaving group on the molecule. Formed intermediate isadditionally stabilized by the hyperconjugation of CF3 groups.Further formation includes the separation of nitrous acid HNO2

(or HNO in the case of nitrosobenzene 3) and phenolate 1a.HNO2 decomposes to give NO2 and NO in the solution, whichmakes the solution dark red. The fact that only one equivalentof Me4NOH is consumed to convert all the nitrobenzene 2 into

romethyl)nitrobenzene 2 and -nitrosobenzene 3 (ref. 16) dissolved inifluoromethyl)phenyl derivative (red spectrum, b) after treatment withmpound formed is pentakis(trifluoromethyl)phenolate 1a.

phenolate 1a by a nucleophile OH�.

This journal is © The Royal Society of Chemistry 2014

Scheme 3 General route from nitrobenzene to phenol 1.

Communication RSC Advances

Publ

ishe

d on

19

Aug

ust 2

014.

Dow

nloa

ded

by U

nive

rsity

of

Tar

tu o

n 22

/09/

2014

10:

14:2

6.

View Article Online

phenolate 1a supports the proposed mechanism shown inScheme 2.

Our initial goal was to obtain pentakis(triuoromethyl)nitrobenzene 2 from pentaiodonitrobenzene,22 a knowncompound, using a triuoromethylation method with pregen-erated CuCF3.13 Obtained nitrobenzene 2 would have been agood starting material for phenol 1, aniline 4 and possibly forsome other polyuorinated substances. As the result of tri-uoromethylation reaction, surprisingly, we did not get penta-kis(triuoromethyl)nitrobenzene 2 but only phenolate 1a.Additional steps are now not necessary to prepare phenol, butunfortunately, the possibility to obtain nitrobenzene 2 in a largescale is also vanished. However, aniline 4 can be obtained easilyfrom chlorobenzene 5 with a yield of 79%. Nitrobenzene 2 canbe obtained from pentakis(triuoromethyl)benzenediazoniumsalt by reacting it with NaNO2, although only at very low yield(13%).16 Several nitrationmethods (nitric acid in the presence ofother acids; AgNO2–BF3 system23) of pentakis(triuoromethyl)benzene 6 were also carried out to achieve respective nitro-benzene 2 but the efforts were unsuccessful most probablybecause of very high positive charge on the ipso-carbon ofcompound 6 and possibly due to steric hindrance by CF3groups. It might be possible to obtain nitrobenzene 2 frompentakis(triuoromethyl)benzene anion but its deprotonationrequires a rather strong base13 and an anhydrous oxygen freereaction environment, where it might be challenging to includea NO2

+ cation.Phenol 1 has been an oen requested compound from other

research groups. The method presented here for obtainingphenol 1 with yield 58% (taking into consideration the purity ofpentaiodonitrobenzene – 57%, vide infra) proceeds via two stepsstarting from commercially available substances such as nitro-benzene, iodine, periodic acid, conc. H2SO4, CuBr, KF, andRuppert–Prakash reagent. The yield is still not very good but theprocess itself is well repeatable, sufficiently pure phenol 1 isobtained (Spectra S8 and S22‡) and the yield is the largest of anyyield published so far. Whenever phenol 1 is required, thismethod published here would be the method of choice.

The rst step – periodination – is carried out in conc. H2SO4

in the presence of iodine and periodic acid as described in theref. 22. Iodination is incomplete and the iodinated productcontains pentaiodo and tetraiodo (two isomers) derivatives.When careful recrystallization was performed (two times fromthe mixture of benzene and ethanol),24 then pure pentaiodo-nitrobenzene with a yield of 22% was obtained. The chro-matographic separation of polyiodinated arenes is not trivialbecause of the scarce solubility and insufficient difference inpolarity to achieve separation. HPLC-ESI-MS-UV-vis analysisshowed that the crude periodinated product prepared by uscontained 31% and 12% tetraiodo derivatives and 57% pen-taiodo derivative (Spectrum S24,‡ UV-vis measurements at252–256 nm). Aer recrystallization, the product contained,according to 1H NMR and HPLC-ESI-MS-UV-vis measure-ments, mainly pentaiodonitrobenzene (Spectrum S25‡). UV-vis peaks were identied according to the signal in ESI nega-tive mode of an adduct of the chlorine anion and periodinatednitrobenzene.

This journal is © The Royal Society of Chemistry 2014

The second step – triuoromethylation – is carried out inthe mixture of NMP and DMF, where CuCF3 is pregeneratedusing CuBr, Ruppert–Prakash reagent and KF. The reactionmixture is stirred overnight; during this time the mixture turnsdark brown, most probably from NO2. Usually it remainslighter brown or even greenish. When the reaction is nished,the solvents and other volatile materials (FSiMe3, less than veCF3-substituted nitrobenzenes, and decomposition products)are removed by vacuum distillation. Phenolate 1a is not vola-tile and remains in the residue, and is extracted from theresidue with diethyl ether. Ether is removed by rotary evapo-ration, and crude phenolate 1a (Spectrum S7‡) is sublimedfrom concentrated sulfuric acid. Sufficiently pure neutralphenol 1 (Spectrum S8‡) is obtained, and no further purica-tion is required (Scheme 3).

In Scheme 4, the hypothetical reaction pathway from nitro-benzene 2 to phenolate 1a during triuoromethylation reactionin DMF and NMP is shown. Instead of nucleophilic OH� species(Scheme 2), basic oxygen from the solvent NMP attacks the mostelectrophilic site of the product.

A blue compound is trapped into the liquid nitrogen trap(Fig. S26‡) when distilling the solvents from the reactionmixture. This blue substance turns gaseous when the trap iswarming, indicating that N2O3 (from NO2 and NO) forms in thereaction. It is an additional proof to the proposed reactionmechanism in Scheme 4. Other by-products (for example 2-bromo-1-methylpyrrole) were not identied.

Reaction between pentaiodochlorobenzene and pregen-erated CuCF3 proceeds in more predicted manner; penta-kis(triuoromethyl)chlorobenzene 5 is obtained in additionto the traces of phenolate 1a and benzene 6. If we would havesimple nucleophilic substitution process, then the chlorineatom, as a better leaving group than the nitro group, wouldbe substituted in the same manner by the oxygen atom.Chlorobenzene 5 is not a very good subject for hydrolysis. Itappears that in addition to electronic effects there are strongconformational effects that forces nitrobenzene 2 intophenolate 1a.

It is known13 that pentakis(triuoromethyl)phenyl deriva-tives are very crowded substances. The most extreme exampleso far is the anion of pentakis(triuoromethyl)phenylmalononitrile C6(CF3)5C(CN)2

�, where the –C(CN)2� group

needs to be planar with the benzene ring to exhibit the bestresonance properties. Because the –C(CN)2

� group is quitelarge and it does not t between ortho-CF3 groups quitewell, then the benzene ring of the compound C6(CF3)5C(CN)2

�

takes obvious bath conformation; the angle between twoplanes of carbon atoms of the benzene ring (one plane: ipso

RSC Adv., 2014, 4, 41895–41901 | 41897

Scheme 4 Possible reaction pathway from nitrobenzene 2 tophenolate 1a by oxygen-nucleophilic solvent NMP during the tri-fluoromethylation reaction.

RSC Advances Communication

Publ

ishe

d on

19

Aug

ust 2

014.

Dow

nloa

ded

by U

nive

rsity

of

Tar

tu o

n 22

/09/

2014

10:

14:2

6.

View Article Online

and ortho carbon atoms; second plane: ortho and meta carbonatoms on the benzene ring) is 35� (!).

A nitro group is not as large as malononitrile but it is largerthan halides. It was not possible to obtain the crystal structureof 2 or 3 because of the extreme volatility and instability of thecrystals, even at low temperatures. Oen, computationalstructure does not differ considerably from the experimentalcrystal structure (Table S20‡ of ref. 13), therefore here weexamine only the computational structure of compounds 2 and3 (pages S4–S6 in ESI‡). When the anionic malononitrile groupforces itself to be planar to the benzene ring, it is different fromnitro- and nitrosobenzene 2 and 3. The nitro group turns itselfout of plane; it is almost transverse to the benzene ring (theangle between the RMS plane of benzene ring and the RMSplane of ipso-C and nitro group is 79.1�), and is therefore inlower resonance with the aromatic system. The nitroso groupalso turns itself out of plane, but because it is slightly smallerand unsymmetrical it is not that transverse to the benzene ring(the angle between the RMS plane of benzene ring and the RMSplane of ipso-C, N and O group is 33.6�), and the benzene ringwith CF3 groups has a denite propeller shape (Fig. 2 and pagesS4–S6 in ESI‡). Because there are no overlapped orbitals thatwould strengthen the bond between the nitro group andbenzene ring in this triuoromethylated system, the distancebetween ipso-C and nitrogen in compounds 2 and 3 are longerthan those for unsubstituted nitrobenzene. The values are1.491 A for unsubstitued nitrobenzene, 1.517 A for 2 and 1.535 A

Fig. 2 Computational geometries of unsubstituted nitrobenzene,pentakis(trifluoromethyl)nitrobenzene 2 and pentakis(trifluoromethyl)nitrosobenzene 3. The compounds are placed in such a manner thatthe benzene ring is horizontal and the position and angle of nitro ornitroso groups are visible.

41898 | RSC Adv., 2014, 4, 41895–41901

for 3. The electronic and conformational synergy in pentakis-(triuoromethyl)nitrobenzene 2 and -nitrosobenzene 3 there-fore makes this system sufficiently reactive to bind to nucleo-philes; for example, nitro and nitroso groups can be detachedmore simply than a chlorine atom in oxygen-nucleophilicenvironment.

In the present work, we can distinguish two different tri-uoromethylation reactions: one with pentaiodonitrobenzenethat contains up to 43% tetraiodonitrobenzenes (hereaerindicated in the text and in ESI as the Reaction I, Spectra S2–S8‡) and the second one with recrystallized pure pentaiodoni-trobenzene (Reaction II, Spectra S9–S14‡).

Obtaining pure pentaiodonitrobenzene is not trivial, recrys-tallization is carried out in rather unpleasant solvents, and therecrystallization yield is low. It is encouraged to carry out thetriuoromethylation reaction with a mixture of pentaiodo- andtetraiodonitrobenzenes, which is obtained from the periodina-tion reaction because tetraiodo derivatives do not give pheno-lates in these conditions and can be removed by vacuumdistillation with the solvents.

Aer the triuoromethylation reaction (Reaction I), theintegrals in 19F NMR spectrum show that the crude reactionmixture contains approximately 45% phenolate 1a, 20%symmetrical nitrobenzene 2,3,5,6-(CF3)4-C6HNO2 and 35% non-symmetrical nitrobenzene 2,3,4,5-(CF3)4-C6HNO2 (SpectrumS2,‡ the position of CF3 groups in the ring can be obtained from19F coupling pattern). HPLC-ESI-MS-UV-vis results of periodi-nated starting materials show the content to be 57, 12, 31%. It ispossible that some of the reactants are reduced during the tri-uoromethylation reaction giving a lower content of phenolate1a and a higher content of tetrakis(triuoromethyl)nitroben-zenes. This similarly occurs when reacting C6I6 with CuCF3: oneof the products in addition to phenolate 1a and C6(CF3)6 wasC6(CF3)5H 6 (14% yield).13 The change in percentages in the caseof this reaction shows that approximately 12% of the pentaio-donitrobenzene is converted into tetrakis(triuoromethyl)nitrobenzene (mainly to the symmetrical one).

To ensure whether tetrakis(triuoromethyl)nitrobenzeneappears during the triuoromethylation process or not, areaction with pure pentaiodonitrobenzene was carried out(Reaction II). Because there were some problems with coolingduring the operation, the reaction did not come out very clean.As the result, the main component according to 19F NMRspectrum was still phenolate 1a but there were approximately50% of other CF3 (Spectrum S9,‡ 19F chemical shi range �45–65 ppm) and C2F5 (�65–90 ppm) containing substances, amongwhich both the isomers of tetrakis(triuoromethyl)nitroben-zenes with contents of up to 30% were present (Spectrum S10‡).They most probably came from the reduction during the tri-uoromethylation process. When the triuoromethylationreaction is carried out with pentaiodochlorobenzene, hydrogensubstitutes the chlorine atom but not iodine; as a by-product weget benzene 6 instead of tetrakis(triuoromethyl)chloroben-zene. The yield of phenol 1 of this specic reaction from pen-taiodonitrobenzene is 45%. The purity according to 19F NMRand GC-MS is very high. It implies that even when extensive

This journal is © The Royal Society of Chemistry 2014

Communication RSC Advances

Publ

ishe

d on

19

Aug

ust 2

014.

Dow

nloa

ded

by U

nive

rsity

of

Tar

tu o

n 22

/09/

2014

10:

14:2

6.

View Article Online

decomposition occurs, it is possible to obtain pure phenol withreasonable yield and high purity.

It would be interesting to know whether tetrakis(tri-uoromethyl)nitrobenzene also gives phenol when reacted withMe4NOH. On adding Me4NOH (25% in MeOH) to the distillatethat contains isomers of tetrakis(triuoromethyl)nitroben-zenes, a dark red solution is formed indicating the formation ofNO2 and NO. Later on, the distillate was further worked up, andthe mixture of tetrakis(triuoromethyl)nitrobenzenes wereseparated (see experimental section). We obtained a solid thatcontained 51% of 2,3,4,5-(CF3)4-C6HNO2 and 49% of 2,3,5,6-(CF3)4-C6HNO2 (according to GC-MS and 19F NMR data, SpectraS6 and S21‡). Then, the mixture of these two substances wasdissolved in MeOH and Me4NOH (25% in MeOH), and wasadded under stirring as long as the drop of the solution did notbecome red in color. MeOH was removed by rotavapor, and theobtained solid was extracted rst with pentane and then withchloroform. Pentane extract contains 14% of symmetrical and86% of non-symmetrical tetrakis(triuoromethyl)phenylderivatives, whose chemical shis now differ from the ppmvalues of initial compounds. The GC-MS revealed that thesecompounds are tetrakis(triuoromethyl)methoxybenzenes(Spectra S16 and S23‡). According to 19F NMR spectrum, chlo-roform extract contains a small amount of pentakis(tri-uoromethyl)phenolate 1a with other uorine containingsubstances (Spectrum S17‡). ESI-HRMS shows the formation oftraces of pentakis- and tetrakis(triuoromethyl)phenolates(Spectrum S20‡). The main quantity of the compound in theresidue from the extractions was found to be tetramethylam-monium nitrate (IR-ATR, Spectrum S18‡).

From this experiment, it is determined that nonsymmetricaltetrakis(triuoromethyl)nitrobenzene is mainly converted intomethoxybenzene, but symmetrical nitrobenzene appears todecompose more extensively, also giving 1a. The pathway to2,3,5,6-(CF3)4-C6HOMe probably goes through formed tetra-methylammonium methoxide, which is an extremely strongnucleophile that reacts with tetrakis(triuoromethyl)nitro-benzene. The solution becomes red or brownish from formedNO2; tetramethylammonium nitrate forms when there is a smallexcess of Me4NOH, nitrous acid, water or oxygen in the solution.

Conclusion

Pentakis(triuoromethyl)nitrobenzene 2 and -nitrosobenzene 3bind in oxygen-nucleophilic environment with an oxygen atomand turn into pentakis(triuoromethyl)phenolate 1a due to theconformational and electronic properties of the molecules.Phenol 1 forms from pentaiodonitrobenzene during the tri-uoromethylation reaction with pregenerated CuCF3 in thesolvents DMF and NMP. NMP is enough oxygen-nucleophilicthat the conversion occurs during the triuoromethylationreaction. Some of the reactants reduce during tri-uoromethylation reaction giving tetrakis(triuoromethyl)nitrobenzene but not pentakis(triuoromethyl)benzene, astypically obtained in triuoromethylation reactions with pen-taiodochlorobenzene or hexaiodobenzene.

This journal is © The Royal Society of Chemistry 2014

Tetraiodonitrobenzenes are converted into tetrakis(tri-uoromethyl)nitrobenzenes during triuoromethylation reac-tion. These do not further react with NMP, do not formphenolates and therefore can be easily separated from theformed phenolate 1a. In the presence of stronger nucleophiles,such as methoxide anion, pentakis(triuoromethyl)nitroben-zenes also undergo a conversion forming either tetrakis(tri-uoromethyl)methoxybenzene or decompose giving traces ofphenolate 1a.

Experimental sectionUsed chemicals

Nitrobenzene (pure) was distilled prior to use. Almost whitecolored (not green as commercial CuBr) CuBr was preparedfrom CuSO4 (pure), KBr (pure) and Na2SO3 (pure) according toref. 25 and dried at 60 �C under vacuum for 6 hours, stored andweighted in a glovebox. H5IO6 (Sigma-Aldrich, $99.0%), I2(99%), conc. H2SO4 (95.0–97.0%), NMP (over Molecular Sieve,according to Karl-Fisher titration, water content 8.5 ppm), DMF(over Molecular Sieve, according to Karl-Fisher titration, watercontent 1.5 ppm), KF (spray-dried, 99%, contains 0.5% SiO2);CF3SiMe3 ($99.0%), Me4NOH (25% in MeOH) and othercommon laboratory solvents were used as received.

Methods

NMR spectra were recorded on a 200 MHz spectrometerequipped with 5mmQNP-z probe for 19F spectra at 188.30 MHz.Chemical shis are referenced to the internal standard C6F6(�162.59 ppm). All the measurements were carried out at 25 �C.HRMS-ESI spectra were recorded on a FT-ICR-MS spectrometer.Samples were dissolved in MeCN or in MeOH, and no buffersolution was used. HPLC-ESI-MS-UV-vis measurements werecarried out by a device using a polar endcapped C18 analyticalcolumn (4.60 mm � 250 mm, 4 mm) with C18 guard cartridge4.0 mm � 2.0 mm. HPLC conditions were as follows: mobilephase A: buffer solution (pH¼ 3.2; 1 mM ammonium acetate in0.1% formic acid); mobile phase B: methanol. The percentage ofB was linearly increased from an initial 40–100% in 20 min,followed by 15 min at 100%. The column was maintained at 30�C and 20 mL of the sample was injected. Samples were dis-solved in MeOH. GC-MS spectra were obtained with 70 eV ofelectron ionization. The GC column was 30m long and 0.25 mmin diameter capillary column. GC program: 40–150 �C: 5 �Cmin�1; 150–200 �C: 10 �C min�1; rst 2.8 minutes was not letinto mass detector. All the samples were dissolved in Et2O (1–2% solution) and 1 mL was injected. Melting points weredetermined with capillary melting point apparatus and areuncorrected. To prevent sublimation during the meltingprocedure the capillaries were sealed before measurements.Computations were carried out by Turbomole 6.5. Geometryoptimization was carried out using def-TVZP basis set. Obtainedcoordinates and values of energy for molecules are presented inESI.‡

RSC Adv., 2014, 4, 41895–41901 | 41899

RSC Advances Communication

Publ

ishe

d on

19

Aug

ust 2

014.

Dow

nloa

ded

by U

nive

rsity

of

Tar

tu o

n 22

/09/

2014

10:

14:2

6.

View Article Online

Pentaiodonitrobenzene C6I5NO2

The reaction was carried out, as reported in ref. 22. Periodic acid(23.85 g, 0.10 mol) was dissolved in conc. sulfuric acid (375 mL).Pounded iodine (78.00 g, 0.30 mol) was added to the clearsolution, and then stirred using an overhead stirrer for half anhour. The reaction mixture was then placed in an ice-bath.Distilled nitrobenzene (7.0 mL, 0.067 mol) was added over 15minutes. The mixture was then stirred for one day at roomtemperature and then for one day at 95 �C. Aer heating thereaction mixture to 95 �C for one day, the iodine sublimed onthe top of the ask. The solution was then carefully poured ontoice. The yellow suspension was then ltered. The residue wasnot washed with ethanol as described in ref. 22 because theproduct is soluble in this solvent. Pentaiodonitrobenzene onthe lter was instead recrystallized using a continuous extractor(Soxhlet extractor may be used as well) from AcOEt (400 mL).One batch of the product was ltered. The volume of the AcOEtwas then reduced to 200 mL and the solution was placed intorefrigerator. The yellow precipitate obtained was ltered again,the two batches were combined, washed with hexane (200 mL)and dried in a vacuum (10�3 Torr) at 50 �C for six hours. 36.4 gof a mixture of penta- (57%) and tetraiodonitrobenzenes (43%)was obtained.

According to the 1H NMR spectrum andHPLC-ESI-MS-UV-vismeasurements, recrystallization twice from benzene andethanol24 gave the substance that has no aromatic hydrogens.The recrystallization yield was 22%.

Pentakis(triuoromethyl)phenol C6(CF3)5OH 1

The reaction was carried out as described for C6I5Cl in ref. 16.Instead of C6I5Cl, C6I5NO2 (may contain C6I4HNO2 derivatives)was used: Dispersion of CuBr (9.71 g, 0.050 mol) and DMF(50 mL) was cooled to 0 �C, triuoromethyltrimethylsilane (9.50mL, 0.067 mol) was added. Aer that, KF (3.86 g, 0.066 mol) wasadded slowly via solid addition funnel. Then, NMP (10 mL) wasincluded, and the reaction mixture was stirred on a ice-bath for1.5–2 hours to form active CuCF3. C6I5NO2 (7.70 g, contains0.006 mol pentaiodonitrobenzene, and 0.005 mol isomers oftetraiodonitrobenzenes) and additional NMP (40 mL) wasadded. The reaction mixture was stirred overnight, and solventswith other volatile substances were distilled off. The residuecontains phenolate 1a, which is further worked up. It isimportant to remove the high boiling solvents as much aspossible from the residue to simplify the following workup. Ithelps if the residue is also washed with pentane to removesolvents but not phenolate 1a. Phenolate 1a was then extractedinto diethyl ether (3 � 70 mL). Ether was evaporated, theobtained solid was transferred into sublimation apparatus,conc. H2SO4 (8 mL) was added and phenol 1 sublimed at60–70 �C at static vacuum (4–5 Torr).

If the residue contains NMP or DMF aer distillation, it isnot possible to solidify the product aer removing ether; an oilysubstance is obtained instead, which has to be transferred intothe sublimation apparatus by dissolving it in a small amount ofether. Prior to the addition of conc. H2SO4 and sublimation,

41900 | RSC Adv., 2014, 4, 41895–41901

ether has to be removed carefully to prevent bubbling of thesolution.

Phenol 1 obtained aer sublimation is slightly hygroscopic;therefore, let the sublimation nger to warm up before scrapingoff the product. Yield from pentaiodonitrobenzene (it has beentaken into account the content of tetraiodo nitrobenzenes fromHPLC-UV-vis measurements – 57%) of phenol 1 is 58% (1.45 g,0.003 mol). M.p. 88–90 �C (lit. 88–90 �C)13,16

19F NMR spectrum of crude reaction mixture (Reaction I,Spectrum S2‡) shows that there are 64% 2,3,4,5-(CF3)4-C6HNO2

and 36% 2,3,5,6-(CF3)4-C6HNO2 (percentages are converted to100% leaving out 1a). Aer distillation, according to 19F NMRspectrum, the ratio is changed to 54 and 46%, respectively(Spectrum S3‡). The content of symmetrical nitrobenzene hasincreased. There were no considerable amounts of 2,3,4,5-(CF3)4-C6HNO2 in the cold-trap. It is possible that some of theCF3 substituted nitrobenzene decomposed during thedistillation.

The distillate was worked up as follows: water (25 mL)was added to the distillate (100 mL) and extracted with pentane(3 � 50 mL). Pentane was removed by rotary evaporation; andthe ask was kept cold to prevent the sublimation ofcompounds. To the obtained oily substance, 2 mL of water wasadded, and crystals formed into this solution. The solution wasltered and the solid on the lter was recrystallized frompentane (25 mL). The obtained solid is the mixture of 2,3,4,5-(CF3)4-C6HNO2 (51%) and 2,3,5,6-(CF3)4-C6HNO2 (49%) (SpectraS6 and S21‡).

Acknowledgements

This study was supported by the Estonian Science Foundation(grants nos 8720 and 8162), the Estonian Research Council(grant no. PUT182), the targeted nancing project of the Esto-nian Ministry of Education and Science (grant no. SF0180089s08), the Estonian Centre of Excellence (HIGH-TECHMATSLOKT117T) and by the Estonian National R & D infrastruc-ture development program of Measure 2.3 “Promotion ofdevelopment activities and innovation” (Regulation no. 34)funded by the Enterprise Estonia foundation.

References

1 V. V. Aksenov, V. M. Vlasov and G. G. Yakobson, J. FluorineChem., 1982, 20, 439–458.

2 M. H. Emmert, A. K. Cook, Y. J. Xie and M. S. Sanford, Angew.Chem., Int. Ed., 2011, 50, 9409–9412.

3 Z. Ke and Y.-Y. Yeung, Org. Lett., 2013, 15, 1906–1909.4 T. Agrawal and S. P. Cook, Org. Lett., 2013, 15, 96–99.5 C. B. Bheeter, J. K. Bera and H. Douceta, Adv. Synth. Catal.,2012, 354, 3533–3538.

6 J. W. W. Chang, E. C. Chia, C. L. L. Chaia and J. Seayad, Org.Biomol. Chem., 2012, 10, 2289–2299.

7 I. Krossing, H. Brands, R. Feuerhake and S. Koenig, J.Fluorine Chem., 2001, 112, 83–90.

8 M. V. Metz, Y. Sun, C. L. Stern and T. J. Marks,Organometallics, 2002, 21, 3691–3702.

This journal is © The Royal Society of Chemistry 2014

Communication RSC Advances

Publ

ishe

d on

19

Aug

ust 2

014.

Dow

nloa

ded

by U

nive

rsity

of

Tar

tu o

n 22

/09/

2014

10:

14:2

6.

View Article Online

9 I. Krossing and I. Raabe, Angew. Chem., Int. Ed., 2004, 43,2066–2090.

10 M. V. Petersen, A. H. Iqbal, L. N. Zakharov, A. L. Rheingoldand L. H. Doerrer, Polyhedron, 2013, 52, 276–283.

11 R. C. Neu, Y. Ouyang, S. J. Geier, X. Zhao, A. Ramos andD. W. Stephan, Dalton Trans., 2010, 39, 4285–4294.

12 N. Ibrahim, M. H. Vilhelmsen, M. Pernpointner,F. Rominger and A. S. K. Hashmi, Organometallics, 2013,32, 2576–2583.

13 A. Kutt, V. Movchun, T. Rodima, T. Dansauer, E. B. Rusanov,I. Leito, I. Kaljurand, J. Koppel, V. Pihl, I. Koppel,G. Ovsjannikov, L. Toom, M. Mishima, M. Medebielle,E. Lork, G.-V. Roschenthaler, I. A. Koppel andA. A. Kolomeitsev, J. Org. Chem., 2008, 73, 2607–2620.

14 A. Kutt, T. Rodima, J. Saame, E. Raamat, V. Maemets,I. Kaljurand, I. A. Koppel, R. Y. Garlyauskayte,Y. L. Yagupolskii, L. M. Yagupolskii, E. Bernhardt,H. Willner and I. Leito, J. Org. Chem., 2011, 76, 391–395.

15 A. Kutt, I. Leito, I. Kaljurand, L. Soovali, V. M. Vlasov,L. M. Yagupolskii and I. A. Koppel, J. Org. Chem., 2006, 71,2829–2838.

This journal is © The Royal Society of Chemistry 2014

16 A. Kutt, F. Werner, I. Kaljurand, I. Leito and I. A. Koppel,ChemPlusChem, 2013, 78, 932–936.

17 Large amount of C6(CF3)5OLi (lithium salt of 1a) wasprepared according to method 2 in the main text forelectrochemical measurements by another research group.No data has been published.

18 S. Roy, B. T. Gregg, G. W. Gribble, V.-D. Le and S. Roy,Tetrahedron, 2011, 67, 2161–2195.

19 H. Morimoto, T. Tsubogo, N. D. Litvinas and J. F. Hartwig,Angew. Chem., Int. Ed., 2011, 50, 3793–3798.

20 E. J. Cho, T. D. Senecal, T. Kinzel, Y. Zhang, D. A. Watson andS. L. Buchwald, Science, 2010, 328, 1679–1681.

21 T. Knauber, F. Arikan, G.-V. Roschenthaler and L. J. Gooßen,Chem.–Eur. J., 2011, 17, 2689–2697.

22 D. L. Mattern, J. Org. Chem., 1984, 49, 3051–3053.23 G. A. Olah, A. P. Fung, S. C. Narang and J. A. Olah, J. Org.

Chem., 1981, 46, 3533–3537.24 G. B. Deacon and G. J. Farquharson, Aust. J. Chem., 1977, 30,

1701–1713.25 L. Hartwell, Org. Synth., 1955, 3, 185–188; L. Hartwell, Org.

Synth., 1944, 24, 22–25.

RSC Adv., 2014, 4, 41895–41901 | 41901