Re

Ta

b

h

•

•

•

•

•

a

ARRAA

KPCNED

1

boup

(

h0

Journal of Hazardous Materials 278 (2014) 491–499

Contents lists available at ScienceDirect

Journal of Hazardous Materials

j o ur nal ho me pa ge: www.elsev ier .com/ locate / jhazmat

eactivity of polychlorinated biphenyls in nucleophilic andlectrophilic substitutions

atyana I. Gorbunovaa,∗, Julia O. Subbotinab, Viktor I. Saloutina, Oleg N. Chupakhina

I. Ya. Postovskii Institute of Organic Synthesis, Ural Branch, Russian Academy of Sciences, Kovalevskoy St., 22, Ekaterinburg 620990, RussiaUral Federal University named after the first President of Russia B.N. Yeltsin, Mira St., 19, Ekaterinburg 620002, Russia

i g h l i g h t s

Quantum chemical calculations werecarried out for PCBs congeners.Calculated descriptors were used toexplain the PCBs reactivity in SN andSE substitutions.Obtained data were used to estimatethe PCBs reactivity in the SN reactions.Calculated descriptors were insuffi-cient to explain the PCBs reactivity inthe SE reactions.New neutralization methods of thelarge-capacity PCBs were discussed.

g r a p h i c a l a b s t r a c t

r t i c l e i n f o

rticle history:eceived 15 January 2014eceived in revised form 9 June 2014ccepted 10 June 2014vailable online 24 June 2014

a b s t r a c t

To explain the chemical reactivity of polychlorinated biphenyls in nucleophilic (SN) and electrophilic(SE) substitutions, quantum chemical calculations were carried out at the B3LYP/6-31G(d) level of theDensity Functional Theory in gas phase. Carbon atomic charges in biphenyl structure were calculated bythe Atoms-in-Molecules method. Chemical hardness and global electrophilicity index parameters weredetermined for congeners. A comparison of calculated descriptors and experimental data for congener

eywords:CBsongenersucleophilic substitutionlectrophilic substitution

reactivity in the SN and SE reactions was made. It is shown that interactions in the SN mechanism arereactions of the hard acid–hard base type, these are the most effective in case of highly chlorinated sub-strates. To explain the congener reactivity in the SE reactions, correct descriptors were not established.The obtained results can be used to carry out chemical transformations of the polychlorinated biphenylsin order to prepare them for microbiological destruction or preservation.

escriptors

. Introduction

The problem of disposal of technogenic polychlorinatediphenyls (PCBs) is still relevant for many countries. The burning

f PCBs is the only technologically developed and effective methodsing a turbojet engine [1]. This method is a highly power-inputrocess as it is necessary to follow the rule of “three T” (reaction

∗ Corresponding author. Tel.: +7 343 369 3058; fax: +7 343 369 3058.E-mail addresses: gorbunova@ios.uran.ru, tigorbunova@yandex.ru

T.I. Gorbunova).

ttp://dx.doi.org/10.1016/j.jhazmat.2014.06.035304-3894/© 2014 Elsevier B.V. All rights reserved.

© 2014 Elsevier B.V. All rights reserved.

time (�) is nearly 2–3 s; reaction temperature (T) is 2000–3000 ◦C;turbulence (t) is very high). The burning of PCBs also requires a largeamount of oxygen (4–6 tons for burning of each PCBs ton). Accord-ing to various data, the amount of PCBs exceeds 1 million tons inthe world. All PCBs cannot be burned. Microbiological and chemicalmethods of PCBs processing are an alternative to burning.

At present microbiological methods of PCBs destruction aregaining importance due to selection of new effective strain

destructors. The most successful studies in the field of the PCBsbiodestruction are those that use lowly chlorinated congeners[2–5]. From their results it was established that the PCBs biodegrad-ability increases with a decrease in the amount of chlorine atoms

4 zardo

iahoe

hrnbfac

iachPBds[c(cealOr

mtgrcpk

etwpetm

2

2

Smgc(

2

da

92 T.I. Gorbunova et al. / Journal of Ha

n the initial substrates. At the same time, there is still consider-ble interest to search for surfactants for solubilization of moreydrophobic highly chlorinated congeners in aqueous medium inrder to create stable emulsions where strains decompose the PCBsffectively [6–8].

Among the chemical methods of PCBs neutralization,ydrodechlorination is developing most intensively [9]. As aule, Pd catalysts have the most positive effect on hydrodechlori-ation. Depending on the Pd catalyst, the source of hydrogen cane both hydrogen generator and hydrogen obtained in situ. In theormer the Pd catalysts modified by inert catalyst carriers, suchs C and SiO2, are applied [10–14], whereas in the latter the Pdatalysts are used with active metals such as Fe and Mg [15–23].

It should be noted that each of the 209 PCBs congeners hasndividual reactivity, and there is no uniform methodologicalpproach to chemical transformations of all the PCBs. This con-lusion concerns commercial mixtures substantially. For example,ydrodechlorination of the decachlorobiphenyl (PCB 209) usingd/C at 209–230 ◦C leads to the formation of biphenyl and benzene.ut the commercial mixture of PCBs Arochlor 1254 in the same con-itions at 130 ◦C is transformed into biphenyl, benzene, toluene,tyrene and propylbenzene mixture [12]. As was shown in Ref.21], the use of Pd/Мg catalyst for the hydrodechlorination of the 4-hlorobiphenyl (PCB 3) and the 2,3,4,5,2′,3′,4′-heptachlorobiphenylPCB 170) in H2O–MeOH medium leads to complete removal ofhlorine atoms from the initial substrates after 10 min. A similarffect for Aroclor 1260 is not observed. Lowly chlorinated PCBsnd biphenyl formed initially are then adsorbed on the Pd cata-yst surface. Following this reaction the process is slowed down.rtho-substituted congeners are the most difficult to hydrodechlo-

inate.In Ref. [24] “selective rules” for hydrodechlorination of com-

ercial mixture Delor 103 were determined. Even for tri- andetrachlorinated congeners from the Delor 103 the authors do notive a definite answer about the advantages of the hydrodechlo-ination of several C Cl bonds. Other commercial mixturesonsisting of more highly chlorinated congeners lead to more com-licated results. In general, for them the selective rules are notnown.

When investigating chemical transformations of PCBs, wencounter problems of an incomplete conversion of congeners andhe absence of reaction selectivity [25]. The purpose of the presentork is to explain the reactivity of the PCBs congeners in nucleo-hilic (SN) and electrophilic (SE) substitutions by means of obtainedxperimental results and quantum chemical non-empirical (ab ini-io) calculations. The results of interactions between commercial

ixture of the PCBs and MeO− (SN) and NO2+ (SE) are considered.

. Experimental

.1. Chemicals

The object of the research is Russian industrial PCBs mixtureovol (analog of commercial mixture Arochlor 1254 (USA)). Sodiumethoxide (MeONa, cp grade, Russia), dimethyl sulfoxide (DMS, cp

rade, Russia), toluene (cp grade, Russia), calcium chloride (CaCl2,p grade, Russia), nitric acid (HNO3) (cp grade, Russia) and oleumH2SO4 (20% SO3), cp grade, Russia) were used.

.2. Chemical experiments

Interaction of PCBs mixture Sovol with sodium methoxide wasescribed [26]. Interaction of PCBs mixture Sovol with nitric acidnd oleum was reported [27].

us Materials 278 (2014) 491–499

2.3. Analytical methods

The PCBs products were analyzed by a GC equipped with a flameionization detector (FID) (Shimadzu, GC-2010) or GC/MS (Agilent7890A MS 5975C Inert XL) [26–28]. The quantitative estimationwas carried out by internal normalization.

2.4. Quantum chemical calculations

Quantum chemical non-empirical calculations were performedusing the Gaussian 09 software package [29]. The structural param-eters were fully optimized in the approximation of the DensityFunctional Theory at the RB3LYP/6-31G(d) level in gas phase. Thereliability of the minimum of the localized stationary points is con-firmed with Hessian’s calculation. For all compounds imaginaryvibrational frequencies are absent.

The used B3LYP/6-31G(d) basis set is small. But literature anal-ysis shows that it yields quite reliable and adequate results inagreement with experimental data for some aromatic substrates[30–34]. There are positive examples of the basis used to calculatethe values for biphenyl derivatives [35] and PCBs [36,37].

To estimate the reactivity of congeners in the mixture Sovolin the SN reaction the following descriptors were chosen: highestoccupied molecular orbital energies (EHOMO), lowest unoccupiedmolecular orbital energies (ELUMO), chemical hardness (�) andglobal electrophilicity index (ω). The two last parameters werecalculated using Eqs. (1)–(5) [36,38–42]:

� = 12

(IP − EA) (1)

IP is the ionization potential and EA is the electron affinity.

IP = −EHOMO (2)

EA = −ELUMO (3)

ω = �2

2�(4)

� is the chemical potential, eV.

� = −12

(IP + EA) (5)

To calculate the charge values (q) on the carbon atoms inthe biphenyl structure Bader’s theory “Atoms-in-Molecules” (AIM)was used [43] as well as the wave function obtained by theRB3LYP/6-31G(d) method. The AIM method was applied to carryout calculations using the AIMAll software package [44].

The basis B3LYP/6-31G(d) used in the present work is very con-venient for q calculations within the Mulliken’s scheme. We carriedout such calculations but the q values were not satisfactory. Forexample, the q values for the carbon atoms at C Cl bonds werenegative but the q values for the carbon atoms at C H bonds werepositive.

The torsion angle between benzene cycles upon rotation withrespect to the simple C C-bond (ϕ) was determined within theframework of the RB3LYP\6-31G(d) method. The scaling factor forthe B3LYP/6-31G(d) level is 0.9945 with respect to the experiment[45].

3. Results and discussion

3.1. Making mixture Sovol

The composition of the commercial mixture Sovol was repre-sented in an earlier paper [25]. It was determined that the Sovol isa mixture of 35 congeners; the contents of tetra-, penta-, and hex-achlorobiphenyls are 22, 56 and 20%, respectively. Minor amounts

T.I. Gorbunova et al. / Journal of Hazardous Materials 278 (2014) 491–499 493

F sible Pf

om

3a

tametoomFttmrfoGcibmtwi

SS

ig. 1. Chromatogram of the mixture Sovol. (The IUPAC designations of all the posont was used.)

f tri- and heptachlorobiphenyls are nearly 1%. A typical chro-atogram of the Sovol is shown in Fig. 1.

.2. Resulting products after interaction between mixture Sovolnd sodium methoxide (SN mechanism)

Interaction of congeners from mixture Sovol with MeONa inhe bipolar aprotic solvent (DMS) proceeds through the SN mech-nism according to Scheme 1. The choice of the interaction as theain one is dictated by the following reasons. There are available

xperimental data for methoxylation of the mixture Sovol [26] andhe individual congeners [28]. The MeONa is the first homologuef alcoholates. It has no considerable steric effect on processingf the nucleophilic aromatic substitution. Earlier identification ofethoxy derivatives was represented by means of GC–MS [26,28].

ig. 2 shows a chromatogram of mixtures obtained from a reac-ion between the mixture Sovol and MeONa. Chemical structures ofhe resulting compounds were determined by scanning every chro-

atographic peak in the GC–MS conditions (70 eV) and subsequenteconstruction of fragmentary ions in the mass spectra. A typicaleature of the mass spectra of methoxy derivatives is the presencef strong molecular peaks for each molecular ion. The obtainedC–MS information makes it possible to estimate the amount ofhlorine atoms and methoxy groups in the methoxy derivatives butt is not sufficient to determine the arrangement of substituents iniphenyl structures owing to a similar fragmentation of various

ethoxy derivatives in the GC–MS conditions. Methoxy deriva-

ives obtained from the heptachlorobiphenyls (PCB 170, PCB 180)ere not detected in the reaction mixture and are not considered

n Scheme 1.

cheme 1. The mixture of the compounds obtained from interaction of the mixtureovol with sodium methoxide.

CBs congeners are given in [50]. To specify numbers of congeners the bold Arabic

From the GC–MS results it is established that all three trichlori-nated congeners (PCB 22, PCB 28, PCB 33) from the mixture Sovolin the reaction with MeONa do not undergo complete conversionand form at least four chemical compounds named monomethoxy-dichlorobiphenyls (5%) (Fig. 2).

The eleven isomer tetrachlorobiphenyls from the mixture Sovolin the reaction with MeONa form mono- (21%, at least thirteenchemical compounds) and dimethoxy derivatives (4%, at least twochemical compounds) (Fig. 2). The three congeners (PCB 44, PCB 49,PCB 52) do not undergo complete conversion. The other tetrachlo-rinated congeners react with MeONa completely.

The thirteen isomer pentachlorobiphenyls from the mixtureSovol after the reaction with MeONa form mono- (24%, at leasttwelve chemical compounds), di- (19%, at least nineteen chemicalcompounds) and trimethoxy derivatives (3%, at least two chemicalcompounds) (Fig. 2).

The mixture Sovol has six hexachlorobiphenyls. After the reac-tion with MeONa they form di- (10%, at least ten chemicalcompounds) and trimethoxy derivatives (5%, at least three chemicalcompounds) (Fig. 2).

The obtained experimental data show that lowly chlorinatedcongeners (tri- and tetrachlorobiphenyls) in the reactions ofnucleophilic aromatic substitution possess lower reactivity incomparison with highly chlorinated congeners (penta- and hex-achlorobiphenyls). Because of various arrangements of chlorineatoms in congeners from the mixture Sovol, from the present exper-imental data it is impossible to determine the reason for PCBsdifferent reactivity. An estimation of the congeners’ reactivity canbe made by means of descriptors obtained from quantum chemicalcalculations.

3.3. Descriptors for estimation of chemical reactivity of PCBscongeners in SN reaction

3.3.1. EHOMO and ELUMO values. Analysis of additional descriptors

The EHOMO and ELUMO values calculated using the AIM method

are presented in Table 1 Their ELUMO decreases from tri- to hex-achlorobiphenyls in most cases, which points to an increase in theelectrophilicity of highly chlorinated molecules in comparison with

494 T.I. Gorbunova et al. / Journal of Hazardous Materials 278 (2014) 491–499

Fig. 2. Chromatogram of methoxy derivatives obtained after interaction between the mixture Sovol and MeONa (two-digit bold Arabic numbers denote PCB congeners that didnot react): 1: monomethoxydichlorobiphenyls (C12H7Cl2(OCH3)); 2: monomethoxytrichlorobiphenyls (C12H6Cl3(OCH3)); 3: dimethoxydichlorobiphenyls (C12H6Cl2(OCH3)2);4 biphed yls (Cb

lcf

chitamu

aatimftgticatp

agP

3

aetto

: monomethoxytetrachlorobiphenyls (C12H5Cl4(OCH3)); 5: dimethoxytrichloroimethoxytetrachlorobiphenyls (C12H4Cl4(OCH3)2); 8: trimethoxytrichlorobiphenold was used.

owly chlorinated congeners. Therefore, the reactivity of the highlyhlorinated PCBs from the mixture Sovol must be higher than thator lowly chlorinated congeners.

Similar conclusions can be made from the analysis of the cal-ulated ω values (Table 1). The ω values increase from tri- toexachlorobiphenyls. It indicates an increase of the electrophilic-

ty as well and it is confirmed by the experimental data [26,28]:ri- and tetrachlorobiphenyls do not undergo complete conversionnd form mono- and dimethoxy derivatives; highly chlorinatedolecules react completely and form more methoxylation prod-

cts.According to Pearson’s acid base concept (HSAB theory) [46]

n effective reaction is the interaction between a soft base and soft acid or between a hard base and a hard acid. Based onhe data that methoxide anion (MeO−) is a hard base [46], itsnteraction with PCBs congeners is more effective if chlorinated

olecules are harder acids. Calculated � indexes for congenersrom the mixture Sovol show that from tri- to hexachlorobiphenylshis parameter increases. Therefore, the highly chlorinated con-eners (tetra-, penta- and hexachlorobiphenyls) are harder acidshan lowly chlorinated molecules, and their interaction with MeO−

s more effective than in the case of trichlorobiphenyls. For suchongeners as PCB 60, PCB 66, PCB 70, PCB 74, PCB 105, PCB 118nd PCB 156 the � values are comparable with the � indexes forrichlorobiphenyls and even lower. However, it is experimentallyroven that their interactions with MeO− are complete [26,28].

The analysis of the calculated values presented in this sectionllows us to carry out a primary estimation of the reactivity of con-ener groups from the mixture Sovol and to predict it for otherCBs.

.3.2. Charges and torsion anglesThe assumption about the interaction between PCBs and MeO−

s the hard acid–hard base type and the calculated values of the

nergy gap (�E) between EHOMO and ELUMO (Table 1) indicate thathese reactions are charge controlled. The high �E values showhat in each PCB molecule there is a clear localization of a surplusf the electronic density on carbon atoms connected with chlorine

nyls (C12H5Cl3(OCH3)2); 6: trimethoxydichlorobiphenyls (C12H5Cl2(OCH3)3); 7:12H4Cl3(OCH3)3). To specify numbers of not reacted congeners the arabic font in

atoms according to the classic Jahn–Teller effect of the first order.Therefore, it is possible to predict the reaction regioselectivity ifaccount is taken of the charge distributions for PCBs molecules.

The q as well as the ϕ values, related to the molecules’ stabilityin the reaction conditions, might also be useful for obtaining addi-tional information about the chemical reactivity. Values of thesedescriptors are presented in Table S1 (see supplementary data).

Supplementary material related to this article can befound, in the online version, at http://dx.doi.org/10.1016/j.jhazmat.2014.06.035.

3.3.2.1. Trichlorobiphenyls. According to our calculations, the car-bon atoms taking part in C Cl bonds in the most substitutedaromatic ring are most deprived of electrons, and MeO− attackswill be directed onto these reactive centers. High chemical reac-tivity of the most substituted aromatic fragments from PCBs wasnoted repeatedly [1,25]. Close q values in the same congener favorsthe absence of selectivity in the SN reactions: for PCB 22 the q val-ues at C-2 and C-3 are equal (0.067 eV); for PCB 28 the q values atC-2 and C-4 are 0.054 eV and 0.056 eV, respectively. It also explainsthe formation of at least four monomethoxy derivatives from threetrichlorobiphenyls.

The calculated ϕ values for trichlorobiphenyls are relatively low(ϕ = 58.4◦, 55.4◦ and 52.2◦ for PCB 22, PCB 28 and PCB 33, respec-tively), which indicates a higher delocalization of the electronicdensity between two aromatic rings. It characterizes PCB 22, PCB 28and PCB 33 as more stable congeners with a low chemical reactivity.It is confirmed by our experimental results [26].

3.3.2.2. Tetrachlorobiphenyls. The asymmetrical tetrachlorinatedcongeners PCB 44 and PCB 49 have higher q values in aromaticunequal rings and rather high ϕ values (ϕ = 83.9◦ and 78.2◦ for PCB44 and PCB 49, respectively). The highest q values in PCB 44 andPCB 49 are observed on carbon atoms in position 2 (q = 0.076 eV

and 0.064 eV for PCB 44 and PCB 49, respectively). Obviously, theattack of MeO− will be directed onto this center.

The symmetrical congener PCB 52 has lower q values(q = 0.053 eV at C-2 and C-2′; q = 0.046 eV at C-5 and C-5′) but a high

T.I. Gorbunova et al. / Journal of Hazardous Materials 278 (2014) 491–499 495

Table 1Calculated values (EHOMO, ELUMO, � and ω) for congeners from the mixture Sovol.

Congeners’ no (indexesof chlorine atoms)

EHOMO (eV) ELUMO (eV) �E (eV) Chemical hardness � (eV) Global electrophilicityindex, ω (eV)

TrichlorobiphenylsPCB 22 (2,3,4′-) −6.580 −1.184 5.396 2.698 2.794PCB 28 (2,4,4′-) −6.513 −1.245 5.268 2.634 2.857PCB 33 (3,4,2′-) −6.564 −1.214 5.350 2.675 2.827

TetrachlorobiphenylsPCB 41 (2,3,4,2′-) −6.897 −0.976 5.921 2.956 2.617PCB 44 (2,3,2′ ,5′-) −6.795 −1.015 5.780 2.890 2.638PCB 47 (2,4,2′ ,4′-) −6.808 −1.208 5.600 2.800 2.868PCB 49 (2,4,2′ ,5′-) −6.811 −1.074 5.737 2.868 2.710PCB 52 (2,5,2′ ,5′-) −6.799 −1.116 5.683 2.842 2.757PCB 56 (2,3,3′ ,4′-) −6.727 −1.375 5.352 2.676 3.066PCB 60 (2,3,4,4′-) −6.639 −1.374 5.265 2.633 3.050PCB 64 (2,3,6,4′-) −6.831 −1.083 5.748 2.874 2.724PCB 66 (2,4,3′ ,4′-) −6.677 −1.428 5.249 2.625 3.128PCB 70 (2,5,3′ ,4′-) −6.718 −1.451 5.267 2.633 3.169PCB 74 (2,4,5,4′-) −6.639 −1.457 5.182 2.591 3.162

PentachlorobiphenylsGroup I

PCB 84 (2,3,6,2′ ,3′-) −6.910 −1.123 5.787 2.893 2.789PCB 91 (2,3,6,2′ ,4′-) −6.950 −1.152 5.798 2.899 2.830PCB 95 (2,3,6,2′ ,5′-) −6.827 −1.163 5.664 2.832 2.818PCB 110 (2,3,6,3′ ,4′-) −6.937 −1.205 5.732 2.866 2.892

Group IIPCB 97 (2,4,5,2′ ,3′-) −6.937 −1.223 5.714 2.857 2.914PCB 99 (2,4,5,2′ ,4′-) −6.913 −1.376 5.537 2.769 3.102PCB 101 (2,4,5,2′ ,5′-) −6.901 −1.290 5.611 2.806 2.990PCB 118 (2,4,5,3′ ,4′-) −6.788 −1.627 5.161 2.580 3.430

Group IIIPCB 82 (2,3,4,2′3′-) −7.007 −1.106 5.901 2.950 2.788PCB 85 (2,3,4,2′ ,4′-) −7.032 −1.153 5.879 2.939 2.850PCB 87 (2,3,4,2′ ,5′-) −6.887 −1.186 5.701 2.851 2.858PCB 105 (2,3,4,3′ ,4′-) −6.801 −1.539 5.262 2.631 3.305PCB 92 (2,3,5,2′ ,5′-) −6.915 −1.246 5.669 2.834 2.937

HexachlorobiphenylsGroup IV

PCB 128(2,3,4,2′ ,3′ ,4′-)

−7.113 −1.298 5.815 2.908 3.042

PCB 132(2,3,4,2′ ,3′ ,6′-)

−7.013 −1.231 5.782 2.891 2.939

PCB 138(2,3,4,2′ ,4′ ,5′-)

−7.025 −1.351 5.674 2.837 3.092

Group VPCB 149(2,3,6,2′ ,4′ ,5′-)

−6.973 −1.270 5.703 2.852 2.980

PCB 153′ ′ ′

−7.075 −1.363 5.712 2.856 3.116

5

ϕasbppo

iCtertpPn

(2,4,5,2 ,4 ,5 -)PCB 156(2,3,4,5,3′ ,4′-)

−6.892 −1.728

value (ϕ = 87.3◦). On the one hand, the presence of two chlorinetoms in ortho-positions results in an instability of the moleculartructure and in an increase of the chemical reactivity of the C Clonds. On the other hand, the absence of chlorine atoms in para-ositions of PCB 52 promotes the stability of the C Cl bonds in SNrocesses and PCB 52 should be rather inactive. It is confirmed byur experimental results [26].

The symmetrical congener PCB 47 shows the opposite behaviorn a reaction with MeONa. The congener PCB 47 has a q value at-2 comparable to that for PCB 52 on the same carbon atom. Theorsion angle for PCB 47, 70.6◦, is lower than that for PCB 52. Appar-ntly, the two chlorine atoms at para-positions for PCB 47 are theeason for complete conversion of this congener. It corresponds

o classical conclusions about beneficial effect of electronegativeara-substituents in SN reactions. Therefore, the charge factor forCB 47 and PCB 52 plays a secondary role in the determination ofucleophilic attacks.

.164 2.582 3.597

Congeners PCB 56, PCB 66 and PCB 70 are characterized by ratherlow ϕ values (ϕ = 58.5◦, 52.5◦ and 52.5◦ for PCB 56, PCB 66 and PCB70, respectively). They have higher q values in aromatic rings withstructural similarities (q = 0.069–0.071 eV). Their reactivity is prac-tically identical. Earlier we found that the individual PCB 70 in thereaction conditions represented in Scheme 1 forms six chemicalcompounds: three monomethoxy and three dimethoxy derivatives[28]. Therefore, the congeners PCB 56 and PCB 66 with similar struc-tures in a reaction with MeONa can be transformed both to monoand dimethoxy derivatives. Monomethoxy derivatives are proba-bly formed upon a nucleophilic attack on the positions 3′ and 4′

and also upon a nucleophilic attack on the carbon atom with thehighest q in a different ring. The presented calculated data do not

allow us to determine the types of dimethoxydichlobiphenyls fromthe congeners PCB 56 and PCB 66.

Congeners PCB 41, PCB 60, PCB 64 and PCB 74 have the high-est q on carbon atom in a trichlorinated ring. The calculated ϕ

4 zardous Materials 278 (2014) 491–499

vPPf(ofPo

3tctPPm8ac

vgct(PsPtb

4agPFnga

tCaacccr

ads

3fr1P

ht(1ta

96 T.I. Gorbunova et al. / Journal of Ha

alues for these congeners show that the most reactive PCBs areCB 41 (ϕ = 88.0◦) and PCB 64 (ϕ = 88.6◦). The highest q values forCB 41 and the PCB 60 are at positions 3 (q = 0.098 eV and 0.097 eVor PCB 41 and PCB 60, respectively), for PCB 64 is at position 2q = 0.081 eV) and for PCB 74 is at position 4 (q = 0.079 eV). With-ut steric effects MeO− should attack these centers. Probably theormation of dimethoxy derivatives for PCB 41, PCB 60, PCB 64 andCB 74 is not excluded as a result of double nucleophilic attacksnto the two carbon atoms with the highest q values.

.3.2.3. Pentachlorobiphenyls. For simplicity, let us divide pen-achlorobiphenyls from the mixture Sovol into groups integratingongeners with the same arrangement of chlorine atoms in one ofhe aromatic rings. Group I consists of PCB 84, PCB 91, PCB 95 andCB 110, they have chlorine atoms at positions 2, 3, 6. CongenersCB 97, PCB 99, PCB 101 and PCB 118 form group II with the arrange-ent of chlorine atoms at positions 2, 4, 5. Group III consists of PCB

2, PCB 85, PCB 87 and PCB 105 with the arrangement of chlorinetoms at positions 2, 3, 4. PCB 92 is separate. It has no structuralommunity with groups I–III.

All of the pentachlorinated congeners from group I have high qalues at C-2 (q = 0.083–0.085 eV). Structural features of PCBs fromroup I result in the highest q values at C-2 if the congener has ahlorine atom at position 2′ and the q value is higher, the closer tohis center is the second chlorine atom. The torsion angles are highϕ = 88.4–91.9◦). The calculated data indicate that PCB 84, PCB 91,CB 95 and PCB 110 are highly reactive substrates. Earlier it washown that in similar conditions the mixture of PCB 105, PCB 107,CB 110 and PCB 118 was transformed in the di- (88.8%) and therimethoxy derivatives (11.2%) but the monomethoxytetrachloro-iphenyls were not formed [28].

All of the congeners from group II have high q values at C- (q = 0.080–0.085 eV). The torsion angles are also high but theyre in general lower than the corresponding parameters for con-eners from group I. It characterizes PCB 97, PCB 99, PCB 101 andCB 118 as less reactive in comparison with those from group I.or all of the congeners from group II the primary center of theucleophilic attack is position 4; position 5 is reactive too. Con-eners from group II can be transformed into both monomethoxynd dimethoxy derivatives [28].

Congeners from group III among all pentachlorobiphenyls havehe highest q values at C-3 (q = 0.098–0.099 eV). Chlorine atoms at-3 are surrounded by ortho- and para-atoms of chlorine. Carbontoms connected with them have high q values too: they are highert C-2 (q = 0.076–0.084 eV) than at C-4 (q = 0.076–0.077 eV). The cal-ulated ϕ values are similar to the corresponding parameters forongeners from group I. For congeners from group III the primaryenter of nucleophilic attack is position 3. Positions 2 and 4 areeactive too.

The calculated q values for PCB 92 are high, they coincide at C-2nd C-3 (q = 0.081 eV). Therefore, the formation of both, mono- andimethoxy derivatives, is possible. The PCB 92 is a very reactiveubstrate, which is also confirmed by a higher ϕ value (ϕ = 84.6◦).

.3.2.4. Hexachlorobiphenyls. Among the hexachlorobiphenylsrom the mixture Sovol congeners with the arrangement of chlo-ine atoms at positions 2, 3, 4 (group IV: PCB 128, PCB 132 and PCB38) can be distinguished. The congeners PCB 149, PCB 153 andCB 156 have no structural community and they form group V.

The hexachlorinated congeners from the mixture Sovolave q values (q = 0.099–0.101 eV), comparable with those forhe pentachlorobiphenyls from group III, and high ϕ values

ϕ = 83.2–91.1◦). It allows us to assume for PCB 128, PCB 132, PCB38, PCB 149 and PCB 153 a high reactivity as in the case of pen-achlorobiphenyls from the group III. The distribution of chlorinetoms in PCB 128, PCB 132, PCB 138, PCB 149 and PCB 153 is

Scheme 2. The mixture of the compounds obtained from interaction of the mixtureSovol with nitric acid and oleum.

the same: there are three chlorine atoms in each aromatic ring.However, in group V there is a congener that has a different struc-ture. It is PCB 156. It possesses the highest q values at positions 3(q = 0.105 eV) and 4 (q = 0.106 eV) and the lowest ϕ values amonghexachlorobiphenyls (ϕ = 59.0◦). In accordance with our experience[28] individual PCB 156 through interaction with MeONa formsmainly monomethoxy derivatives (62.9%) but they were not foundafter the interaction between the mixture Sovol and MeONa [26].Obviously, the low content of PCB 156 in the mixture Sovol and,probably, joint elution of some methoxy derivatives do not make itpossible to determine correctly the types of methoxylation prod-ucts from the PCB 156.

Thus, the analysis of the q values and ϕ values allows us to estab-lish additional features of the reactivity of PCBs congeners and toestimate possible directions of the primary nucleophilic attack.

3.3.3. Experimental NMR 13C spectra for PCBs congenersNMR 13C spectra provide information about the correctness of

the calculated q values. Their analysis consists in comparing q val-ues with the lowest fields of chemical shifts of the carbon nucleiconnected with chlorine atoms. For example, in Refs. [47,48] it wasdetermined that for PCB 28 the nucleus at C-4 has the lowest fieldsof chemical shifts. From our calculations it follows that PCB 28 hasthe highest q value at C-4. High convergence by the criterion “thehighest positive q value at the carbon atom connected with a chlorineatom – the lowest fields of chemical shifts in NMR 13C spectrum” isalso shown by trichlorinated PCB 33 [48], the tetrachlorinated PCB52 [49] and PCB 70 [47], the pentachlorinated PCB 101 [47,49] andthe hexachlorinated PCB 153 [47,49].

3.4. Resulting products after nitration of mixture Sovol (SE

mechanism)

The interaction of congeners from the mixture Sovol with nitricacid and oleum mixture proceeds through the SE mechanism inaccordance with Scheme 2 [27].

Fig. 3 shows a chromatogram of the products obtained afternitration. Chemical structures of the resulting compounds wereestablished by scanning every chromatographic peak in the GC–MSconditions (70 eV) and subsequent reconstruction of fragmentaryions in mass spectra. A typical feature of the mass spectra of nitroderivatives is the presence of strong molecular peaks of each com-pound. As in the case of methoxy derivatives, it is not possibleto index nitro substituents in biphenyl structure. Nitro derivativesfrom the trichlorobiphenyls (PCB 22, PCB 28, PCB 33) and the hep-tachlorobiphenyls (PCB 170, PCB 180) were not detected in thereaction mixture and are not considered in Scheme 2.

Unlike the reaction with MeONa (SN mechanism), the inter-action of the mixture Sovol with nitric acid and oleum mix (SEmechanism) undergoes complete conversion. No initial congenersin the resulting products were found.

From the GC–MS results it was found that the eleven isomerictetrachlorobiphenyls from the mixture Sovol form dinitro deriva-

tives at nitration (16%, at least nine chemical compounds) (Fig. 3).

The thirteen isomeric pentachlorobiphenyls from the mix-ture Sovol transform into mononitro (9%, at least six chemical

T.I. Gorbunova et al. / Journal of Hazardous Materials 278 (2014) 491–499 497

Fig. 3. Chromatogram of nitro derivatives: 1: dinitrotetrachlorobiphenyls (C12H4Cl4(NO2)2); 2: mononitropentachlorobiphenyls (C12H4Cl5NO2); 3: dinitropentachloro-b initro

cc

cc

3S

pectpp

ah1corttTfoP1c

bpratbe

iphenyls (C12H3Cl5(NO2)2); 4: mononitrohexachlorobiphenyls (C12H3Cl6NO2); 5: d

ompounds) and dinitro derivatives (51%, at least thirteen chemicalompounds) (Fig. 3).

The six hexachlorobiphenyls form mononitro (12%, at least ninehemical compounds) and dinitro derivatives (7%, at least fivehemical compounds) (Fig. 3).

.5. Choice of descriptors for estimation of congener reactivity inE reactions

At first sight, the conclusions on nitration results are very sim-le. The deciding factor seems to be the steric factor, which easilyxplains the direction of nitronium cation’s (NO2

+) attack on thearbon atom in the C H bond. Here it is necessary to considerhat chlorine atoms in the SE reactions are orientants at ortho- andara-positions [50], and electrophilic attacks will be carried outreferably onto the carbon atom with the highest negative q.

However, it is a simplified view on the problems of the SE mech-nism. Among the PCBs from the mixture Sovol there are congenersaving no negative q values (PCB 52, PCB 91, PCB 92, PCB 95, PCB01, PCB 132, PCB 149, PCB 153) (see Table S1). It is also necessary toonsider that SN reactions proceed at the most substituted ring and,n the contrary, SE reactions have affinity to most non-substituteding having electron-donating properties. Can it be suggested thathe hexachlorinated PCB 156 upon nitration forms dinitro deriva-ives only on one from the most non-substituted aromatic rings?he answer is ambiguous since all the calculated ϕ values are farrom 0◦ and all the structures are non-planar. It leads to an increasef the chemical reactivity including SE reactions. Such congeners asCB 56, PCB 60, PCB 66, PCB 70, PCB 74, PCB 118, PCB 105 and PCB56 possess rather small ϕ values but they underwent completeonversion upon nitration.

If one considers the Pearson’s theory and proves the nitrationy interaction between hard acid (NO2

+) and hard base (PCB), unex-lained will be the fact that the tetrachlorinated biphenyls in thiseaction do not form mononitro derivatives, although their � values

re similar to those for some penta- and hexachlorobiphenyls fromhe mixture Sovol. To designate nitration of PCBs as the interactionetween soft acid (NO2

+) and soft base (PCB) would not be correctither since a nitronium cation is not a soft acid.

hexachlorobiphenyls (C12H2Cl6(NO2)2).

It is known that a limiting stage in SE reactions for aromaticcompounds is the formation of a transition state (�-complex)[51,52]. For this reason, it appears necessary to calculate parame-ters for them (EHOMO, ELUMO, Fukui indexes, free energy, localizationenthalpy, etc.). On the other hand, such quantum chemical calcu-lations are very complicated since polyaromatic structures of PCBsand �-complexes can be formed into both aromatic rings. Until nowmany researchers have performed quantum chemical calculationsof nitro derivatives [50,53–55]. These data cannot be adapted toPCBs because of their polyaromaticity, and a single aromatic ringcannot be considered to be a simple substituent or orientant of acertain type. For example, the biphenyl in nitration forms 2- and4-nitrobiphenyls [56]. Here the second phenyl ring is orientantat ortho- and para-positions but it is not an axiom for PCBs withchlorine substituents in both rings. It might be necessary to formu-late special approaches within the DFT theory with respect to PCBsmolecules. They are expected to help explain the reactivity of PCBscongeners in SE processes.

Thus, the calculated data cannot explain the congener reactivityat nitration. The descriptors � and ω do not clarify the situation.It is not possible to explain the primary attack of NO2

+ on carbonatoms by means of charge distributions. For instance, among hex-achlorobiphenyls from the mixture Sovol (see Table S1) there isonly one PCB 156 with a single negative charge on the carbon atom.A different approach is needed instead of the AIM scheme. Chargecalculations according to the Mulliken’s scheme appear to be evenmore unreasonable. The calculations result in negative charges ofthe carbon atoms in C Cl bonds and in positive charges in C Hbonds. The Mulliken’s scheme is suitable neither for SE reactions,nor for SN processes.

At present, there is no clarity in the choice of a descriptor toestimate the reactivity of PCBs from the reactionary mixture in SEreactions.

3.6. About practical application of obtained results

The present work is focused on commercial PCBs. They arelarge-capacity technogenic waste and are at a storage stage.Their temporary storage can be suspended in connection with

4 zardo

rra

cbm[nptdlto

oatPP

4

thwawiPoctceTabcvo

tpsc

tomPf

A

R

R

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[of polychlorobiphenyls congeners in the nucleophilic substitution reactions,

98 T.I. Gorbunova et al. / Journal of Ha

equirements of the Stockholm Convention (2001). The obtainedesults can serve as an alternative of the neutralization of toxic PCBsnd as a stage of their utilization by microbiological methods.

Summarizing the results of the presented research, we con-lude that chemical interaction through the SN mechanism cane carried out using highly chlorinated congeners and their com-ercial mixtures, for example, Arochlor 1260 or Clophen A60

57]. Arochlor 1260 consists of penta-, hexa-, hepta-, octa- andonachlorobiphenyls. Clophen A60 consists of tetra- (nearly 1%),enta-, hexa-, hepta- and octachlorobiphenyls. For the neutraliza-ion of toxic PCBs the realization of the SN processes leads to aecrease of chlorine content. It will positively influence the toxicity

evel of the initial substrates and the availability of obtained deriva-ives for strain-destructors due to a decrease of the hydrophobicityf the resulting products.

Chemical interaction through the SE mechanism can be carriedut using all types of PCBs since decachlorobiphenyl (PCB 209) isbsent in commercial mixtures. These derivatives can be preservedemporarily if they are solid. In our research we have not shownCBs sulfonation, however, organic salts from sulfo derivatives ofCBs can be made in the solid state. They can be preserved too.

. Conclusions

From the results of the presented research it was determinedhat interactions between PCBs congeners in SN reactions are of theard acid–hard base type. Chemical hardness (�) of PCBs increasesith an increase of chlorine atoms in the substrates. It confirms

higher reactivity of highly chlorinated congeners in comparisonith that of lowly chlorinated PCBs. For each used nucleophil an

ndividual approach is necessary for predicting the reactivity ofCBs since the nucleophil changes its hardness. Reactive directionf nucleophilic substitution in an aromatic ring is controlled by theharge distribution. It is caused by a large energy gap (�E) betweenhe molecular orbitals (HOMO and LUMO). Charge value (q) at thearbon atoms bonded with chlorine atom is the main descriptor forstimating the direction of the attack onto non-steric nucleophil.he preference of the nucleophilic attack onto more substitutedromatic rings is explained by an increase of the q value on the car-on atom. Therefore, its electrophilicity (ω) increases from lowlyhlorinated to highly chlorinated congeners. The closeness of the qalues at the carbon atoms in the C Cl bonds explains the absencef the selectivity in the SN reactions of PCBs.

Reliable descriptors for estimating PCBs reactivity in SE reac-ions have not been established yet. It may be suggested that therimary electrophilic attack can be carried out onto the least sub-tituted aromatic ring of PCBs in which there are electron-donatingarbon atoms.

Since PCBs reactions proceeding through the SN mechanism leado a decrease of chlorine atoms in the initial substrates, they aref utmost importance for the preparation of technogenic PCBs foricrobiological destruction. The significance of the SE reactions for

CBs can be determined by the synthesis of solid products availableor preservation.

cknowledgment

This research was supported financially by the Ural Branch ofussian Academy of Sciences (project no 12-М-34-2036).

eferences

[1] L.N. Zanaveskin, V.A. Averyanov, Polychlorobiphenyls: problems of the pollu-

tion of the environment and technological neutralisation methods, Russ. Chem.Rev. 67 (1998) 713–724.

[2] L.A. Beaudette, S. Davies, P.M. Fedorak, O.P. Ward, M.A. Pickard, Comparisonof gas chromatography and mineralization experiments for measuring loss

[

us Materials 278 (2014) 491–499

of selected polychlorinated biphenyl congeners in cultures of white rot fungi,Appl. Environ. Microbiol. 64 (1998) 2020–2025.

[3] A. Kubátová, P. Erbanová, I. Eichlerová, L. Homolka, F. Nerud, V. Sasek, PCBcongener selective biodegradation by the white rot fungus Pleurotus ostreatusin contaminated soil, Chemosphere 43 (2001) 207–215.

[4] M.O. Ilori, G.K. Robinson, S.A. Adebusoye, Degradation and mineralization of2-chloro-, 3-chloro- and 4-chlorobiphenyl by a newly characterized naturalbacterial strain isolated from an electrical transformer fluid-contaminated soil,J. Environ. Sci. 20 (2008) 1250–1257.

[5] D.O. Egorova, V.A. Demakov, E.G. Plotnikova, Bioaugmentation of a polychloro-biphenyl contaminated soil with two aerobic bacterial strains, J. Hazard. Mater.261 (2013) 378–386.

[6] R. Fiebig, D. Schulze, J.C. Chung, S.T. Lee, Biodegradation of polychlorinatedbiphenyls (PCBs) in the presence of a bioemulsifier produced on sunflower oil,Biodegradation 8 (1997) 67–75.

[7] K.A. Billingsley, S.M. Backus, S. Wilson, A. Singh, O.P. Ward, Remediation of PCBsin soil by surfactant washing and biodegradation in the wash by Pseudomonassp. LB400, Biotechnol. Lett. 24 (2002) 1827–1832.

[8] N.G. Rojas-Avelizapa, R. Rodriguez-Vazquez, S. Saval-Bohorquez, P.J.J. Alvarez,Effect of C/N/P ratio and nonionic surfactants on polychlorinated biphenylbiodegradation, World J. Microb. Biot. 16 (2000) 319–324.

[9] B.Z. Wu, H.Y. Chen, S.J. Wang, C.M. Wai, W. Liao, K.H. Chiu, Reductive dechlori-nation for remediation of polychlorinated biphenyls, Chemosphere 88 (2012)757–768.

10] Y. Noma, M. Ohno, S.I. Sakai, Pathways for the degradation of PCBs bypalladium-catalyzed dechlorination, Fresen. Environ. Bull. 12 (2003) 302–308.

11] J.G. Rodriguez, A. Lafuente, A new advanced method for heterogeneous cat-alyzed dechlorination of polychlorinated biphenyls (PCBs) in hydrocarbonsolvent, Tetrahedron Lett. 43 (2002) 9581–9583.

12] E. Lorenc-Grabowska, J. Yperman, G. Gryglewicz, S. Hoste, R. Carieer, Studyon hydrodechlorination of polychlorinated biphenyls during pyrolysis in thepresence of a catalyst, Fuel 85 (2006) 374–381.

13] A. Kume, Y. Monguchi, K. Hattori, H. Nagase, H. Sajiki, Pd/C-catalyzed practicaldegradation of PCBs at room temperature, Appl. Catal. B: Environ. 81 (2008)274–282.

14] M. Aresta, A. Dibenedetto, C. Fragale, P. Gianoccaro, C. Pastore, D. Zammiello, C.Ferragina, Thermal desorption of polychlorobiphenyls from contaminated soilsand their hydrodechlorination using Pd- and Rh-supported catalysts, Chemo-sphere 70 (2008) 1052–1058.

15] N.E. Korte, O.R. West, B.G. Liang, J.L. Zutman, Q. Fernando, The effect of solventconcentration on the use of palladized-iron for the step-wise dechlorination ofpolychlorinated biphenyls in soil extracts, Waster Manage. 22 (2002) 343–349.

16] G.V. Lowry, K.M. Johnson, Congener-specific dechlorination of dissolved PCBsby microscale and nanoscale zerovalent iron in a water/methanol solution,Environ. Sci. Technol. 38 (2004) 5208–5216.

17] Y. Fang, S.R. Al-Abed, Dechlorination kinetics of monochlorobiphenyls by Fe/Pd:effects of solvents, temperature, and PCB concentration, Appl. Catal. B: Environ.78 (2008) 371–380.

18] K. Venkatachalam, X. Arzuaga, N. Chopra, V.G. Gavalas, J. Xu, D. Bhattacharyya,B. Henning, L.G. Bachas, Reductive dechlorination 3,3′ ,4,4′-tetrachlorobiphenyl(PCB77) using palladium or palladium/iron nanoparticles and assessment of thereduction in toxic potency in vascular endothelial cells, J. Hazard. Mater. 159(2008) 483–491.

19] M.D. Engelman, R. Hutchenson, K. Henschied, R. Neal, I.F. Cheng, Simultaneousdetermination of total biphenyl and dichlorodiphenyltrichloroethane (DDT) bydechlorination with Fe/Pd and Mg/Pd bimetallic particles and flame ionizationdetection gas chromatography, Microchem. J. 74 (2003) 19–25.

20] S. Agarwal, S.R. Al-Abed, D.D. Dionysiou, Enhanced corrosion-based Pd/Mgbimetallic system for dechlorination of PCBs, Environ. Sci. Technol. 41 (2007)3722–3727.

21] E. Hadnagy, L.M. Rauch, K.H. Gardner, Dechlorination of polychlorinatedbiphenyls, naphthalenes and dibenzo-p-dioxins by magnesium/palladiumbimetallic particles, J. Environ. Sci. Health A 42 (2007) 685–695.

22] O. Wu, W.D. Marshall, Approaches to the determination of polychlori-nated biphenyl (PCB) concentrations in soils/sediments by dechlorination tobiphenyl, Int. J. Environ. Anal. Chem. 80 (2001) 27–38.

23] S. Ishihara, A. Ido, Y. Monguchi, H. Nagase, H. Sajiki, Pd/C-catalyzed dechlo-rination of polychlorinated biphenyls under hydrogen gas-free conditions, J.Hazard. Mater. 229–230 (2012) 15–19.

24] J. Hetflejs, M. Czakóová, R. Rericha, J. Vcelák, Catalyzed dehalogenation of Delor103 by sodium hydridoaluminate, Chemosphere 44 (2001) 1521–1529.

25] T.I. Gorbunova, V.I. Saloutin, O.N. Chupakhin, Chemical methods of transforma-tion of polychlorobiphenyls, Russ. Chem. Rev. 79 (2010) 511–530.

26] O.N. Zabelina, T.I. Gorbunova, M.G. Pervova, V.E. Kirichenko, A.Ya. Zapevalov,V.I. Saloutin, O.N. Chupakhin, Reactivity of congeners of Sovol technical mixtureof polychlorinated biphenyls toward sodium methoxide, Russ. J. Appl. Chem.77 (2004) 1523–1527.

27] T.I. Gorbunova, A.Ya. Zapevalov, V.E. Kirichenko, V.I. Saloutin, O.N. Chupakhin,Preparation of amino derivatives from industrial mixture of polychloro-biphenyls, Russ. J. Appl. Chem. 74 (2001) 118–122.

28] T.I. Gorbunova, M.G. Pervova, V.I. Saloutin, O.N. Chupakhin, Reactivity features

Russ. J. Gen. Chem. 82 (2012) 138–143.29] M.J. Frisch, G.W. Trucks, H.B. Schlegel, G.E. Scuseria, M.A. Robb, J.R. Cheeseman,

G. Scalmani, V. Barone, B. Mennucci, G.A. Petersson, H. Nakatsuji, M. Caricato,X. Li, H.P. Hratchian, A.F. Izmaylov, J. Bloino, G. Zheng, J.L. Sonnenberg, M. Hada,

zardo

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

T.I. Gorbunova et al. / Journal of Ha

M. Ehara, K. Toyota, R. Fukuda, J. Hasegawa, M. Ishida, T. Nakajima, Y. Honda, O.Kitao, H. Nakai, T. Vreven, J.A. Montgomery, J.E. Peralta, F. Ogliaro, M. Bearpark,J.J. Heyd, E. Brothers, K.N. Kudin, V.N. Staroverov, R. Kobayashi, J. Normand, K.Raghavachari, A. Rendell, J.C. Burant, S.S. Iyengar, J. Tomasi, M. Cossi, N. Rega,J.M. Millam, M. Klene, J.E. Knox, J.B. Cross, V. Bakken, C. Adamo, J. Jaramillo,R. Gomperts, R.E. Stratmann, O. Yazyev, A.J. Austin, R. Cammi, C. Pomelli, J.W.Ochterski, R.L. Martin, K. Morokuma, V.G. Zakrzewski, G.A. Voth, P. Salvador,J.J. Dannenberg, S. Dapprich, A.D. Daniels, Ö. Farkas, J.B. Foresman, J.V. Ortiz, J.Cioslowski, D.J. Fox, Gaussian 09, Revision C.01, Gaussian, Inc., Wallingford, CT,2009.

30] N.J.R. van Eikema Hommes, T. Clark, Regression formulae for ab initio anddensity functional calculated chemical shifts, J. Mol. Model. 11 (2005) 175–185.

31] S. Qiu, X. Tan, K. Wu, A. Zhang, S. Han, L. Wang, Experimental and the-oretical study on molecular structure and FT-IR, Raman, NMR spectra of4,4′-dibromodiphenyl ether, Spectrochim. Acta A 76 (2010) 429–434.

32] M. Arivazhagan, R. Meenakshi, Quantum chemical studies on structure of 1-3-dibromo-5-chlorobenzene, Spectrochim. Acta A 82 (2011) 316–326.

33] V. Krishnakumar, D. Barathi, R. Mathammal, Molecular structure, vibrationalspectra, HOMO, LUMO and NMR studies of 1,2-dichloro-4-nitrobenzene and2,3,5,6-tetrachloro-1-nitrobenzene based on density functional calculations,Spectrochim. Acta A 86 (2012) 196–204.

34] V. Krishna Kumar, S. Suganya, R. Mathammal, Molecular structure, vibrationalspectra, HOMO, LUMO and NMR studies of 2,3,4,5,6-penta bromo toluene andbromo durene based on density functional calculations, Spectrochim. Acta A125 (2014) 201–210.

35] F. Ceccacci, G. Mancini, P. Mencarelli, C. Villani, Determination of the rotationalbarrier of a chiral biphenyl: comparison of theoretical and experimental data,Tetrahedron: Asymmetr. 14 (2003) 3117–3122.

36] R. Parthasarathi, J. Padmanabhan, V. Subramanian, B. Maiti, P.K. Chattaraj,Chemical reactivity profiles of two selected polychlorinated biphenyls, J. Phys.Chem. A 107 (2003) 10346–10352.

37] J. Padmanabhan, R. Parthasarathi, V. Subramanian, P.K. Chattaraj, Chemicalreactivity analysis on 3,3′ ,4,4′ ,5,5′-hexa chlorobiphenyl: a DFT approach, J. Mol.Struct.: Theochem. 730 (2005) 221–226.

38] R.G. Parr, W. Yang, Density Functional Theory of Atoms and Molecules, OxfordUniversity Press, New York, 1989.

39] R. Parr, L. Szentpaly, S. Liu, Electrophilicity index, J. Am. Chem. Soc. 121 (1999)1922–1924.

40] P.K. Chattaraj, B. Maiti, U. Sarkar, Philicity: a unified treatment of chemicalreactivity and selectivity, J. Phys. Chem. A 107 (2003) 4973–4975.

41] E. Chamorro, M. Duque-Norena, R. Notario, P. Pérez, Intrinsic relative scales ofelectrophilicity and nucleophilicity, J. Phys. Chem. A 117 (2013) 2636–2643.

42] Z. Zhou, R.G. Parr, Activation hardness: new index for describing the orien-tation of electrophilic aromatic substitution, J. Am. Chem. Soc. 112 (1990)5720–5724.

[

us Materials 278 (2014) 491–499 499

43] R.F.W. Bader, Atoms in Molecules: A Quantum Theory, Oxford University Press,UK, 1994.

44] T.A. Keith, AIMAll (Version 13.02.26), TK Gristmill Software, Overland Park, KS,USA, 2012 (aim.tkgristmill.com).

45] A.P. Scott, L. Radom, Harmonic vibrational frequencies: an evaluation ofHartree–Fock, Møller–Plesset, quadratic configuration interaction, DensityFunctional Theory, and semiempirical scale factors, J. Phys. Chem. 100 (1996)16502–16513.

46] R.G. Pearson, J. Songstad, Application of the principle of hard and soft acids andbases to organic chemistry, J. Am. Chem. Soc. 89 (1967) 1827–1836.

47] M. Yanagisawa, K. Hayamizu, O. Yamamoto, 13C chemical shifts of chlorinatedbiphenyls, Magn. Reson. Chem. 24 (1986) 1013–1014.

48] H.J. Lehmler, L.W. Robertson, Synthesis of polychlorinated biphenyls (PCBs)using the Suzuki-coupling, Chemosphere 45 (2001) 137–143.

49] N.K. Wilson, Carbon-13 nuclear magnetic resonance. 13C shieldings and spin-lattice relaxation times in chlorinated biphenyls, J. Am. Chem. Soc. 97 (1975)3573–3578.

50] M. Ge, L. Yao, Substituent effect on electronic structures of halonitrobenzenes,Spectrochim. Acta A 71 (2008) 1499–1502.

51] A. Arrieta, F.P. Cossıo, Loss of aromaticity and p-electron delocalization in thefirst step of the electrophilic aromatic nitration of benzene, phenol and ben-zonitrile, J. Mol. Struct.: Theochem. 811 (2007) 19–26.

52] A.N. Pankratov, Electrophilic aromatic substitution regioselectivity for benzenederivatives in terms of cationic localization energies from semiempirical quan-tum chemical computations, J. Mol. Struct.: Theochem. 507 (2000) 239–244.

53] P.C. Chen, Y.C. Chieh, Density functional theory study of the internal rotationalbarriers of some aromatic nitro compounds, J. Mol. Struct.: Theochem. 583(2002) 173–180.

54] V. Krishnakumar, R. Sangeetha, R. Mathammal, D. Barathi, Density func-tional theory, comparative vibrational spectroscopic studies, HOMO–LUMO,first hyperpolarizability analyses of 2-fluoro 5-nitrotoluene and 2-bromo 5-nitrotoluene, Spectrochim. Acta A 104 (2013) 77–86.

55] M. Govindarajan, M. Karabacak, A. Suvitha, S. Periandy, FT-IR, FT-Raman, abinitio, HF and DFT studies, NBO, HOMO–LUMO and electronic structure calcu-lations on 4-chloro-3-nitrotoluene, Spectrochim. Acta A 89 (2012) 137–148.

56] R.B. Moodie, K. Schofield, A.R. Wait, Electrophilic aromatic substitution. Part31. l. The kinetics and products of nitration of naphthalene, biphenyl, and somereactive monocyclic aromatic substrates in aqueous phosphoric acid containingnitric acid or propyl nitrate, J. Chem. Soc. Perkin Trans. II (1984) 921–926.

57] K. Ballschmiter, M. Zell, Analysis of polychlorinated biphenyls (PCB) by glass

capillary gas chromatography, Fresen. Zeitsch. Anal. Chem. 302 (1980) 20–31.

58] O.V. Dorofeeva, V.P. Novikov, N.F. Moiseeva, V.S. Yungman, Density functionaltheory study of conformations, barriers to internal rotations and torsionalpotentials of polychlorinated biphenyls, J. Mol. Struct.: Theochem. 637 (2003)137–153.