Formation of Trichlorinated Dibenzo-p-dioxins from 2,4-Dichlorophenol and 2,4,5-Trichlorophenolate: A Theoretical Study

DOI: 10.1002/cphc.200600355

Formation of Trichlorinated Dibenzo-p-dioxinsfrom 2,4-Dichlorophenol and 2,4,5-Trichlorophenolate: A Theoretical StudyErnesto Surez, Dimas Surez, Maria Isabel Men�ndez, Ram�n L�pez,* andToms Luis Sordo[a]

Introduction

Polychlorinated dibenzo-p-dioxins (PCDDs, see Scheme 1) havebecome a subject of great interest because some of their iso-mers are considered as highly toxic compounds in the environ-

ment.[1–4] Most of the emissions of PCDDs stem from anthropo-genic activities such as waste incineration,[5–11] metallurgicalprocesses,[12–16] chemical manufacturing, or the papermakingprocess.[17] To a small extent, PCDDs are also discharged intothe environment from natural processes by volcanic eruptionsand forest and brush fires,[18–20] or even by plants, insects, bac-teria, etc.[21–25] Consequently, PCDDs have been detected inmany areas of our biosphere, such as in the atmosphere,rivers, lakes, seas, and soils.[24,26, 27]

Many experimental studies have focused on the formationof PCDDs in incineration processes.[28–39] In general, the mostimportant mechanisms of formation start from organic precur-

sors like chlorophenols, chlorobenzenes, etc. These processesare usually catalyzed by metals or metal oxides, which arepresent in the surface of fly ashes, and take place at tempera-tures between 550 and 750 K. In the absence of catalysts,these reactions can also occur in the gas phase at tempera-tures of 800–1000 K. An experimental gas-phase study hasshown that the conversion of trichlorophenols into PCDDs in-creases with temperature when going from 600 to 900 K.[30]

The proposed mechanisms for the generation of PCDDsfrom chlorophenols imply, in a first step, the coupling of neu-tral species with chlorophenoxy radicals or chlorophenolateanions stabilized by resonance to give predioxins, which inturn become dioxaspiro-type intermediates that evolvethrough a Smiles rearrangement to afford PCDDs. It is believedthat the last step, the ring closure for PCDD formation, is oneof the key steps in all the mechanisms from organic precursors.The anionic mechanism has been invoked in experimental

[a] E. Su�rez, Dr. D. Su�rez, Dr. M. I. Men�ndez, Dr. R. L�pez, Prof. T. L. SordoDepartamento de Qu�mica F�sica y Anal�ticaFacultad de Qu�mica, Universidad de OviedoC/Juli�n Claver�a 8, 33006 Oviedo, Principado de Asturias (Spain)Fax: (+34)985-10-3125E-mail : [email protected]

Supporting information for this article is available on the WWW underhttp://www.chemphyschem.org or from the author.

The reaction of the 2,4,5-trichlorophenolate anion with 2,4-di-chlorophenol to afford trichlorinated dibenzo-p-dioxins (T3CDDs)is investigated at the B3LYP/6-31+G(d) and B3LYP/6-311+G-ACHTUNGTRENNUNG(3df,2p)//B3LYP/6-31+G(d)+ZPVE(B3LYP/6-31+G(d)) levels oftheory. The first stage of the process corresponds to the forma-tion of a predioxin, which can evolve through four differentroutes. Two of them lead directly to the products 2,3,7-T3CDDand 1,3,8-T3CDD, and the other two afford different predioxin-type intermediates, which in turn can evolve through all or someof the four routes to give new predioxins or T3CDD. Consequent-ly, the theoretical results obtained show plainly the complexchemistry implied in the formation of dioxins from chlorophenolsvia anionic mechanisms by disclosing all the critical structuresthrough which the system evolves, thus allowing assessment of

the viability of the different mechanistic routes and the accessibleproducts. The statistical thermodynamics treatment at 1 atm and298.15, 600, 900, and 1200 K indicates that at higher tempera-tures, the Gibbs energy barrier for the formation of the initial pre-dioxin is clearly the rate-determining step for the whole process,but at lower temperatures the Gibbs energy barrier for this stepis similar to those for its evolution into 2,3,7-T3CDD. This result isin contrast with previous proposals that the closure of the centralring is the rate-limiting step. Finally, according to our results therate constant for the formation of polychlorinated dibenzo-p-di-oxins increases with the temperature, in agreement with the ex-perimental observation that the conversion of trichlorophenolsincreases when going from 600 to 900 K in the gas phase in theabsence of catalysts, and with DFT molecular dynamics results.

Scheme 1. General structure of polychlorinated dibenzo-p-dioxins (PCDDs).

ChemPhysChem 2006, 7, 2331 – 2338 > 2006 Wiley-VCH Verlag GmbH&Co. KGaA, Weinheim 2331

studies in waste incineration[29,30] and in soils of forests for ex-plaining the obtained dioxins.[24] As an illustration, in Scheme 2we display the proposed mechanism for the reaction of the2,4,5-trichlorophenolate anion (2,4,5-T3CPA) with 2,4-dichloro-phenol (2,4-D2CP).

It is difficult to obtain experimental results related to the for-mation pathways of PCDDs due to the toxicity of these com-pounds and the need to achieve experimental conditions simi-lar to the real ones. Therefore, theoretical studies can be rele-vant for understanding the mechanisms of PCDD formation incombustion processes. Most of the theoretical studies have fo-cused on the determination of thermodynamic proper-ties,[4, 35,40–48] while the remainder have analyzed the moleculargeometry,[49–51] vibrational frequencies,[50] mechanisms of for-mation[35,52–55] and decomposition,[56] and reactivity with OHradicals.[57] The mechanisms of formation of PCDDs involvinganionic species have been studied less than those involvingradicals and focusing only on the closure step of the centralring for PCDD formation. An Austin Model 1 (AM1) semiempiri-cal study on the reaction between 2,4,6-trichlorophenol and2,4,6-T3CPA has indicated the existence of a transition state(TS) with an energy barrier of 13.7 kcalmol�1 for the passagefrom a predioxin to a dioxaspiro-type intermediate, which inturn evolves into PCDD formation via Cl� elimination withoutencountering any activation energy.[52] A molecular dynamicsstudy employing DFT methodology has also investigated thesame process at temperatures of 600 and 900 K starting fromfour different predioxins.[55] The most favorable reaction pathstarts from the same predioxin as in the above-mentionedsemiempirical study, and the reaction rate increases with thetemperature when passing from 600 to 900 K.

With the aim of obtaining detailed knowledge of the anionicreaction mechanisms that afford PCDDs in the gas phase inthe absence of catalysts, we have undertaken a DFT study of

the reaction of 2,4,5-T3CPA with 2,4-D2CP. We have also ana-lyzed the effect of the temperature on the studied reactiveprocess to explain some experimental facts.

Results and Discussion

First, we present the different mechanisms of the formation ofPCDDs for the reaction of 2,4,5-T3CPA with 2,4-D2CP. Then, wediscuss the effect of temperature on these results for compari-son with the experimental findings.

Reaction Mechanisms

Table 1 shows the relative B3LYP/6-31+G(d) and B3LYP/6-311+GACHTUNGTRENNUNG(3df,2p)//B3LYP/6-31+G(d)+ZPVE(B3LYP/6-31+G(d))energies with respect to separate reactants, and the zero-pointvibrational energies (ZPVEs) of the most significant structures

involved in the formation of PCDDs from the reactants used inthis investigation. Figure 1 displays the corresponding opti-mized geometries. Unless otherwise stated, we cite in the textthe relative B3LYP/6-311+G ACHTUNGTRENNUNG(3df,2p)//B3LYP/6-31+G(d)+ZPVE(B3LYP/6-31+G(d)) energies.

The first stage of the reactive process studied is an aromaticnucleophilic substitution (see Figure 1), in which the anionicO atom of 2,4,5-T3CPA attacks the chlorinated C atom in theortho position of 2,4-D2CP to give an intermediate I1, 20.1 kcalmol�1 below the reactants, through the TS tsR-I1, 26.7 kcalmol�1 above the reactants. At this TS the attacking O atom is1.787 J from the attacked C atom, and the salient Cl atom sep-arates 2.073 J from the C atom to which it was initiallybonded. At I1 a bridged O atom links the two phenyl ringsand the salient Cl atom migrates towards the hydroxyl H atomat a distance of 1.938 J. This intermediate evolves to give pre-

Scheme 2. Proposed dioxaspiro pathway for the formation of 2,3,7-trichloro-dibenzo-p-dioxin (2,3,7-T3CDD).

Table 1. Relative B3LYP/6-31+G(d) electronic energies (DE), ZPVEs, andrelative B3LYP/6-311+G ACHTUNGTRENNUNG(3df,2p)//B3LYP/6-31+G(d) electronic energies in-cluding ZPVEs (D ACHTUNGTRENNUNG(E’+ZPVE)) [kcalmol�1] of the most significant chemicalstructures located for the reaction between 2,4,5-T3CPA and 2,4-D2CP.

Species DE ZPVE D ACHTUNGTRENNUNG(E’+ZPVE)

2,4,5-T3CPA+2,4-D2CP 0.0 92.9 0.0tsR-I1 25.4 92.6 26.7I1 �23.2 93.9 �20.1A+HCl 4.7 90.3 2.8tsA-a+HCl 26.5 89.7 24.5tsA-b+HCl 22.3 89.5 20.1tsA-c+HCl 73.4 87.3 68.1tsA-d+HCl 95.6 86.3 89.1B+HCl 3.0 90.2 0.8tsB-a+HCl 26.0 89.7 24.0tsB-c+HCl 74.1 87.3 68.9tsB-d+HCl 94.2 86.3 87.7C+HCl 0.7 90.4 �0.72,3,7-T3CDD+HCl+Cl- 1.7 90.9 3.11,3,8-T3CDD+HCl+Cl- 1.3 91.0 2.81,3,7-T3CDD+HCl+Cl- 1.3 91.1 3.1

2332 www.chemphyschem.org > 2006 Wiley-VCH Verlag GmbH&Co. KGaA, Weinheim ChemPhysChem 2006, 7, 2331 – 2338

R. L�pez et al.

www.chemphyschem.org

dioxin A, 2.8 kcalmol�1 above the reactants, after eliminatingan HCl molecule without any TS. As a consequence, theO atom of the monochlorinated ring now supports the biggestpart of the negative charge of the system, and the distance be-tween this O atom and the substituted C atom in the ortho po-sition of the monochlorinated ring (1.269 J) is clearly shorterthan the corresponding one at tsR-I1 (1.380 J).

Predioxin A can proceed through four different pathways viafour TSs (see Figure 1). Pathway (a) is direct Cl� elimination byan intramolecular aromatic nucleophilic substitution whereby

the nonbridged O atom attacks the chlorinated C atom, whichis in the ortho position with respect to the bridged O atom, togive the products 2,3,7-trichlorodibenzo-p-dioxin (2,3,7-T3CDD)+Cl� , 3.1 kcalmol�1 higher in energy than the separatereactants. The TS for this step is tsA-a, 24.5 kcalmol�1 abovethe reactants; the distance between the attacking O atom andthe attacked chlorinated C atom is 1.842 J and the salientCl atom is separated 1.891 J from the C atom to which it wasoriginally linked. Pathway (b) corresponds to an intramoleculararomatic nucleophilic substitution in which the nonbridged

Figure 1. B3LYP/6-31+G(d) optimized geometries of the most chemically significant structures located for the reaction of 2,4,5-T3CPA with 2,4-D2CP to giveT3CDDs. Only the most relevant distances (in J) are displayed.

ChemPhysChem 2006, 7, 2331 – 2338 > 2006 Wiley-VCH Verlag GmbH&Co. KGaA, Weinheim www.chemphyschem.org 2333

Formation of Trichlorinated Dibenzo-p-dioxins


O atom attacks the C atom linked to the bridged O atom. It isan isomerization step via a Smiles rearrangement whose TStsA-b, 20.1 kcalmol�1 above the reactants, presents the twophenyl rings in a perpendicular position by means of twobridged O atoms. This TS leads to another predioxin B,0.8 kcalmol�1 above the reactants, in which the chlorinatedC atom of the monochlorinated phenyl moiety is in a para po-sition with respect to the bridged O atom. Pathway (c) is theintramolecular nucleophilic attack of the nonbridged O atomon the unsubstituted C atom of the trichlorinated phenylmoiety in the ortho position with respect to the bridgedO atom, with a simultaneous 1,4-H shift to give a new prediox-in C, 0.7 kcalmol�1 under the reactants. At C the nonbridgedO atom is placed again in the trichlorinated phenyl moiety. Atthe TS for this step, tsA-c (68.1 kcalmol�1 above the reactants),the forming C�H bond is 1.455 J long while the breaking oneis 1.515 J. Pathway (d) consists of an intramolecular nucleo-philic attack of the nonbridged O atom on the nonchlorinatedC atom in the ortho position with respect to the bridgedO atom for a Cl� elimination to give the products 1,3,8-T3CDD+Cl� , 2.8 kcalmol�1 higher in energy than the separate

reactants. The TS for this step is tsA-d, 89.1 kcalmol�1 abovethe reactants, in which the distance between the nonbridgedO atom and the attacked C atom is 2.541 J and the salientCl atom is separated by 1.804 J from the original C atom.Given that H� is not a good salient group, its elimination isavoided by a 1,2-H shift to eliminate the closest Cl� ion.

Therefore, from predioxin A four basic mechanisms havebeen located. Pathways (a) and (d) yield two isomers ofT3CDD, and (b) and (c) give different predioxin intermediatesthat can undergo all or some of the processes correspondingto the basic mechanisms to render new T3CDD or predioxins,which in turn can follow the basic mechanisms and so on (seeFigure 2). Specifically, predioxin B, for example, could evolvethrough the four basic pathways (a)–(d) and return to prediox-in A through route (b). Figure 1 shows the TSs for the possibleevolution of predioxin B. As could be expected, these TSs arevery similar to those for predioxin A both in geometry and inrelative energy, the main difference being the position of theCl atom in the least substituted ring. The question that arisesis: how many and which T3CDDs could be obtained if the in-termediates formed follow the four basic mechanisms?

Figure 2. Schematic view of all the possible intermediates and products that can be formed following the routes found in this work.


R. L�pez et al.


Figure 2 shows all the predioxins and T3CDDs which can begenerated from the reaction of 2,4,5-T3CPA and 2,4-D2CP. The14 predioxin intermediates that appear in Figure 2 can inter-convert along pathway (b) (through dioxaspiro species) orpathway (c) (1,4-H shift). Predioxins G and H can undergo twodifferent 1,4-H shifts because they present two nonequivalentH atoms ortho to the bridge O atom. This fact allows the diver-sification of the reaction paths depicted in Figure 2. On theother hand, predioxins C, D, G, and H cannot evolve throughpathway (a) as they do not have a Cl atom suitable for directelimination, and predioxins C, D, K, and L cannot followroute (d) because a 1,2-H shift does not produce Cl� elimina-tion. The scheme shown in Figure 2, therefore, is useful for un-derstanding the very complex chemistry involved in the reac-tion between 2,4,5-T3CPA and 2,4-D2CP, and shows all the tet-rachlorinated predioxins and T3CDDs that in principle could beobtained.

An alternative way to obtain the same products shown inFigure 2 could be from the reaction of the 2,4-dichloropheno-late anion (2,4-D2CPA) with 2,4,5-trichlorophenol (2,4,5-T3CP).The instability of the TS for the attack of the anionic oxygenatom on the chlorinated carbon atom in the ortho position of2,4,5-T3CP relative to 2,4-D2CPA+2,4,5-T3CP is now6.1 kcalmol�1 less than that of tsR-I1 with respect to 2,4-D2CP+2,4,5-T3CPA. However, 2,4-D2CPA and 2,4,5-T3CP areenergetically 5.7 kcalmol�1 less stable than the separate reac-tants considered in the present work (see Supporting Informa-tion). Therefore, the global rate-determining energy barriers forboth reaction channels, which also imply the same chemistry,would be practically equal as well.

Gibbs Energy Profiles: Effect of Temperature

To assess the kinetic and thermodynamic viability of the stud-ied process, we carried out a statistical thermodynamics treat-ment of the electronic energy profiles described above.Figure 3 displays the DG profiles at 1 atm and 298.15 K for thefour basic mechanisms from predioxin A. Our results indicatethat all the structures along the energy profiles are less stablein relative Gibbs energies than in relative electronic energiesexcept for the separate products. Due to the entropic factor,this destabilization is more accentuated for the first stage ofthe process (tsR-I1 and I1) and consequently tsR-I1 presentsthe highest Gibbs energy barrier (39.3 kcalmol�1), although itis not much larger than those corresponding to the two (a)-type routes leading to 2,3,7-T3CDD from predioxin A. As seenin Figure 3, pathways (c) and (d) are quite unfavorable becausethey present very high energy barriers. Therefore, predioxin Ais expected to evolve along routes (a) and (b). Route (a) with aGibbs energy barrier of 37.9 kcalmol�1 directly yields 2,3,7-T3CDD+Cl� , while route (b) renders predioxin B, which in turnevolves through an (a)-type pathway to give the same trichlori-nated dioxin with a Gibbs energy barrier of 37.4 kcalmol�1.These two different evolution routes of predioxin A presentsimilar energy barriers both corresponding to (a)-type routes.As a consequence, 2,3,7-T3CDD will be the product obtained,in good agreement with the experimental proposal for the re-action of 2,4,5-T3CPA with 2,4-D2CP.[24]

For comparison with experimental results obtained at differ-ent temperatures in the gas phase in the absence of cata-lysts,[30] we also investigated the influence of temperature at

Figure 3. Gibbs energy profiles at 298.15 K and 1 atm of the most important routes involved in the reaction between 2,4,5-T3CPA and 2,4-D2CP.




600, 900, and 1200 K on the DG profiles. Table 2 presents theDH, �TDS, and DG values of the most significant structures in-volved in the studied process.

An increase of temperature from 600 to 1200 K destabilizesagain all of the structures along the energy profiles except forthe products, and particularly tsR-I1 and I1; thus A+HCl nowbecomes more stable than I1 at 900 and 1200 K (see Figure 4).

Therefore, the first step is clearly the rate-determining one andnot the step corresponding to the formation of the tricyclicring, as had previously been indicated.[52,55] According to ourresults, the Gibbs energy barrier increases when passing from298.15 K (39.3 kcalmol�1) to 1200 K (74.3 kcalmol�1) but thisrise is overcompensated by the increase of temperature, andconsequently the rate constant for the formation of 2,3,7-T3CDD will increase. Taking into account the transition-statetheory rate constant equation at two different temperatures[(Eq. (1)] ,

kðT 0ÞkðTÞ ¼ T 0

T

� �exp �DG0�

RT 0 þ DG�

RT

� �ðT 0 > TÞ ð1Þ

the rate constant of formation of 2,3,7-T3CDD at 1200 K isabout 7.6N1015 times greater than that at 298.15 K. However,it is interesting to note that this increase of rate constant di-minishes with the increase of temperature. The ratio of rateconstants at 900 and 600 K is about 3870, while at 1200 and900 K it is about 74. The value of 3870 is in accordance withthe rise of conversion of trichlorophenols into T3CDDs be-tween 600 and 900 K that was found experimentally in the gasphase in the absence of catalysts,[30] as well as with moleculardynamics results.[55]

Conclusions

The theoretical study of the formation of T3CDDs by the reac-tion of 2,4,5-T3CPA with 2,4-D2CP at the B3LYP/6-31+G(d) andB3LYP/6-311+G ACHTUNGTRENNUNG(3df,2p)//B3LYP/6-31+G(d)+ZPVE(B3LYP/6-31+G(d)) levels of theory indicates that the first stage of theprocess corresponds to nucleophilic attack of the oxygen atomof the anion on the neutral molecule followed by eliminationof an HCl molecule to form a predioxin. This intermediate canevolve through four different routes: a) direct elimination ofCl� , b) isomerization through a dioxaspiro-type structure, c) a1,4-H shift, and d) a 1,2-H shift with a simultaneous Cl� elimina-tion. The pathways (a) and (d) lead directly to the products2,3,7-T3CDD+Cl� and 1,3,8-T3CDD+Cl� , respectively, while (b)and (c) afford predioxin-type intermediates, which in turn canevolve through all or some of the routes mentioned above.The routes implying H shifts, (c) and (d), are notably less stablethan the others. The theoretical results obtained in the presentwork show plainly the complex chemistry implied in the forma-tion of dioxins from chlorophenols via anionic mechanisms bydisclosing all the critical structures through which the systemevolves, thus allowing assessment of the viability of the differ-ent mechanistic routes and the accessible products.

The statistical thermodynamics analysis at 1 atm and at298.15, 600, 900, and 1200 K reveals that all the species exceptfor the products destabilize with an increase in temperaturedue to the entropic factor. This effect is particularly importantin the first TS for the formation of the initial predioxin. At

Table 2. DH, �TDS, and DG values in kcalmol�1 of the most significant structures involved in the reaction between 2,4,5-T3CPA and 2,4-D2CP at the tem-peratures of 600, 900, and 1200 K and a pressure of 1 atm.

Species 600 K 900 K 1200 KDH �TDS DG DH �TDS DG DH �TDS DG

2,4,5-T3CPA+2,4-D2CP 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0tsR-I1 27.8 23.5 51.3 28.5 34.4 62.9 29.1 45.2 74.3I1 �19.1 20.9 1.8 -18.1 30.1 12.0 �17.0 38.9 21.9A+HCl 4.4 3.1 7.5 4.8 4.2 9.0 5.3 5.0 10.3tsA-a+HCl 25.1 7.9 33.0 25.0 12.0 37.0 25.0 16.0 41.0tsA-b+HCl 20.8 7.5 28.3 20.8 11.2 32.0 20.8 15.0 35.8tsA-c+HCl 68.7 8.6 77.3 69.0 12.6 81.6 69.2 16.6 85.8tsA-d+HCl 90.6 4.1 94.7 91.0 5.6 96.6 91.2 7.2 98.4B+HCl 2.4 3.4 5.8 2.9 4.5 7.4 3.4 5.4 8.8tsB-a+HCl 24.5 7.9 32.4 24.5 11.9 36.4 24.4 16.0 40.4tsB-c+HCl 69.5 8.5 78.0 69.7 12.5 82.2 69.9 16.4 86.3tsB-d+HCl 89.2 4.1 93.3 89.6 5.6 95.2 89.8 7.3 97.12,3,7-T3CDD+HCl+Cl- 4.2 �9.9 �5.7 4.2 �14.8 �10.6 4.3 �19.9 �15.6

Figure 4. Variation of the most favorable Gibbs energy profiles with temper-ature for the formation of 2,3,7-T3CDD from the reaction between 2,4,5-T3CPA and 2,4-D2CP. Gibbs energy profiles at 298.15 K are included as a ref-erence.


R. L�pez et al.


lower temperatures the energy barrier for the formation of thispredioxin is similar to those for its evolution into 2,3,7-T3CDD.However, at higher temperatures the energy barrier for thefirst step is clearly the rate-determining one for the whole pro-cess. This finding is in contrast with previous proposals thatthe closure of the central ring is the rate-limiting step. Accord-ing to our results, the rate constant for the formation ofT3CDDs increases with temperature, in agreement with the ex-perimentally observed rise of the conversion of trichlorophe-nols when going from 600 to 900 K in the gas phase in the ab-sence of catalysts, and with DFT molecular dynamics results.

Computational Methods

Quantum chemical calculations were performed with the Gaussi-an 98 series of programs.[58] Full geometry optimizations of stablespecies and TSs located were carried out with the B3LYP DFT[59–61]

method and the 6-31+G(d) basis set[62] using the Schlegel algo-rithm.[63] The nature of the stationary points was further checkedand ZPVEs were evaluated by analytical computations of harmonicvibrational frequencies. Intrinsic reaction coordinate (IRC) calcula-tions were also carried out at the B3LYP/6-31+G(d) level to checkthe connection between some of the most significant critical struc-tures located using the Gonzalez and Schlegel method implement-ed in Gaussian 98.[64,65] To obtain a higher energy level, B3LYP/6-311+GACHTUNGTRENNUNG(3df,2p) single-point calculations[66,67] were also performedon the B3LYP/6-31+G(d) optimized geometries, which is conven-tionally denoted as B3LYP/6-311+GACHTUNGTRENNUNG(3df,2p)//B3LYP/6-31+G(d)(B3LYP/6-311+GACHTUNGTRENNUNG(3df,2p)//B3LYP/6-31+G(d) plus ZPVE(B3LYP/6-31+G(d)) if ZPVE is included in the energy evaluation).

DH, DS, and DG values were also calculated to obtain results morereadily comparable with experiment within the ideal gas, rigidrotor, and harmonic oscillator approximations.[68] A pressure of1 atm and temperatures of 298.15, 600, 900, and 1200 K were as-sumed in the calculations.

To check the reliability of our computational scheme, we first opti-mized the full geometry of 2,3,7,8-tetrachlorodibenzo-p-dioxin atthe B3LYP/6-31+G(d) level of theory, and compared it with X-raydiffraction experimental values.[69] As shown in Figure 5, the com-puted bond lengths and bond angles differed from the experimen-tal values by +0.07 to +1.16% (7.33% for the C�H bond) and�0.74 to +0.43%, respectively. These deviations are similar tothose corresponding to previous theoretical calculations,[53,55] thus

confirming the validity of our computational method for moleculargeometry. To further assess the performance of the B3LYP/6-311+GACHTUNGTRENNUNG(3df,2p)//B3LYP/6-31+G(d) level of theory as compared with abinitio methodologies, we also studied the aromatic nucleophilicsubstitution reaction between a phenolate anion and a chloroben-zene molecule to give a diphenyl ether molecule and a chlorideanion (see Scheme 3). The molecular geometries of the reactants,

TS, and products were then optimized at the MP2/6-31+G(d) andB3LYP/6-31+G(d) levels of theory followed by analytic frequencycalculations. Single-point B3LYP energy calculations were per-formed on the B3LYP/6-31+G(d) structures using the 6-311+G-ACHTUNGTRENNUNG(3df,2p) basis set. To estimate the effect of larger basis sets andmore elaborate N-electron treatments on the ab initio energies,CCSD(T)/6-31+G(d) and MP2/6-311+GACHTUNGTRENNUNG(3df,2p) single-point calcu-lations on the MP2/6-31+G(d) geometries were also performed.The corresponding ab initio Gibbs energies were obtained usingan additive combination of electronic energies (i.e. E ACHTUNGTRENNUNG(CCSD(T)/6-31+G*)+E(MP2/6-311+GACHTUNGTRENNUNG(3df,2p))�E(MP2/6-31+G*)) and MP2/6-31+G* thermal corrections. The B3LYP and MP2 methods predicta quite similar molecular geometry for the TS of the model reac-tion between chlorobenzene and phenolate (see Supporting Infor-mation). The computed Gibbs energy barrier in the gas phaseamounts to 33.4 and 25.3 kcalmol�1 according to the best B3LYP-based and ab initio results, respectively. In terms of the reactionGibbs energy, the B3LYP value (�10.0 kcalmol�1) is very close tothe ab initio value (�11.9 kcalmol�1). These results indicate thatthe B3LYP method gives accurate thermochemical data and mayoverestimate the energy barrier of the aromatic nucleophilic sub-stitution processes by about 8 kcalmol�1 with respect to ab initiomethods. Taking into account that the energetic differencesamong the competing reaction paths found in our study are largerthan 8 kcalmol�1, we conclude that the B3LYP mechanistic predic-tions should be reliable.

Acknowledgments

We are grateful to the Ministerio de Ciencia y Tecnolog�a (Spain)for financial support (project number PPQ2001-3442-C02-01).

Keywords: chlorophenols · density functional calculations ·dioxins · environmental chemistry · reaction mechanisms

[1] G. Choudary, L. H. Keith, C. Rappe, Chlorinated Dioxins and Dibenzofur-ans in the Total Environment, Butterworth Publishers, Boston, 1983.

[2] C. Rappe, Chemosphere 1992, 25, 41.

Figure 5. B3LYP/6-31+G(d) optimized geometry of 2,3,7,8-tetrachlorodiben-zo-p-dioxin. Bond lengths and bond angles are given in angstroms and de-grees, respectively. Experimental values[69] are also shown in parentheses asa reference.

Scheme 3. Model reaction studied with both ab initio and B3LYP methodol-ogies. Gibbs energy barrier (DG*) and reaction Gibbs energy (DGrxn) at298.15 K and 1 atm are in kcalmol�1.




[3] C. Rappe, Chemosphere 1993, 27, 211.[4] O. V. Dorofeeva, V. S. Iorish, N. F. Moiseeva, J. Chem. Eng. Data 1999, 44,

516.[5] K. Olie, P. L. Vermeulen, O. Hutzinger, Chemosphere 1977, 6, 455.[6] R. R. Bumb, W. B. Crummett, S. S. Cutie, J. R. Gledhill, R. H. Hummel,

R. O. Kagel, L. L. Lamparski, E. V. Luoma, D. L. Miller, T. J. Nestrick, L. A.Shadoff, R. H. Stehl, J. S. Woods, Science 1980, 210, 385.

[7] F. W. Karasek, L. C. Dickson, Science 1987, 237, 754.[8] R. Batcher, M. Swerev, B. Ballschmiter, Environ. Sci. Technol. 1992, 26,

1649.[9] L. C. Dickson, D. Lenoir, O. Hutzinger, Environ. Sci. Technol. 1992, 26,

1822.[10] R. Addink, H. A. J. Govers, K. Olie, Environ. Sci. Technol. 1998, 32, 1888.[11] A. Yasuhara, T. Katami, T. Okuda, N. Ohno, T. Shibamoto, Environ. Sci.

Technol. 2001, 35, 1373.[12] S. J. Buckland, D. J. Hannah, J. A. Taucher, R. W. Allison, Organohalogen

Compd. 1990, 3, 219.[13] C. Rappe, Organohalogen Compd. 1992, 9, 5.[14] U. Lahl, Chemosphere 1994, 29, 1939.[15] A. Buekens, L. Stieglitz, K. Hell, H. Huang, P. Segers, Chemosphere 2001,

42, 729.[16] C. Xhrouet, E. de Pauw, Environ. Sci. Technol. 2004, 38, 4222.[17] C. Naf, D. Broman, H. Pettersen, C. Rolff, Y. Zebuhr, Environ. Sci. Technol.

1992, 26, 1444.[18] E. D. Jong, J. A. Field, H.-E. Spinnler, J. B. P. A. Wijnberg, J. A. M. D. Bont,

Appl. Environ. Microbiol. 1994, 60, 264.[19] G. W. Gribble, Environ. Sci. Technol. 1994, 28, 310A.[20] P. RuokojTrvi, M. Ettala, P. Rahkonen, J. Tarhanen, J. Ruuskanen, Chemo-

sphere 1995, 30, 1697.[21] T. Eisner, L. B. Hendry, D. B. Peakall, J. Meinwald, Science 1971, 172, 277.[22] L. G. Uberg, B. Glas, S. E. Swanson, C. Rappe, K. G. Paul, Arch. Environ.

Contam. Toxicol. 1990, 19, 930.[23] L. G. Uberg, N. Wagman, R. Andersson, C. Rappe, Organohalogen

Compd. 1993, 11, 297.[24] E. J. Hoekstra, H. de Weerd, E. W. B. de Leer, U. A. T. Brinkman, Environ.

Sci. Technol. 1999, 33, 2543.[25] N. J. L. Green, A. Hassanin, A. E. Johnston, K. C. Jones, Environ. Sci. Tech-

nol. 2004, 38, 715.[26] M. L. Hitchman, R. A. Spackman, N. C. Ross, C. Agra, Chem. Soc. Rev.

1995, 423.[27] E. Eljarrat, J. Caixach, J. Rivera, Environ. Sci. Technol. 2001, 35, 3589.[28] K. Ballschmiter, I. Braunmiller, R. Niemczyk, M. Swerev, Chemosphere

1988, 17, 995.[29] J. G. P. Born, P. Mulder, R. Louw, Environ. Sci. Technol. 1993, 27, 1849.[30] R. Luijk, D. M. Akkerman, P. Slot, K. Olie, F. Kapteijn, Environ. Sci. Technol.

1994, 28, 312.[31] H. Huang, A. Buekens, Chemosphere 1995, 31, 4099.[32] R. Addink, K. Olie, Environ. Sci. Technol. 1995, 29, 1425.[33] K. Tuppurainen, I. Halonen, P. RuokojTrvi, Chemosphere 1998, 36, 1493.[34] J.-Y. Ryu, J. A. Mullholand, Combust. Sci. Technol. 2001, 169, 107.[35] J. A. Mullholand, U. Akki, W. Yang, J.-Y. Ryu, Chemosphere 2001, 42, 719.[36] T. Katami, A. Yasuhara, T. Okuda, T. Shibamoto, Environ. Sci. Technol.

2002, 36, 1320.[37] R. Louw, S. I. Ahonkhai, Chemosphere 2002, 46, 1273.

[38] L. Khachatryan, A. Burcat, B. Dellinger, Combust. Flame 2003, 132, 406.[39] C. S. Evans, B. Dellinger, Environ. Sci. Technol. 2003, 37, 1325.[40] W. M. Shaub, Thermochim. Acta 1982, 58, 11.[41] W. M. Shaub, Thermochim. Acta 1982, 55, 59.[42] J. F. Unsworth, H. Dorans, Chemosphere 1993, 27, 351.[43] D. Thompson, Chemosphere 1994, 29, 2545.[44] C.-L. Huang, B. K. Harrison, J. Madura, J. Dolfing, Environ. Toxicol. Chem.

1996, 15, 824.[45] N. Saito, A. Fuwa, Chemosphere 2000, 40, 131.[46] L. A. Le�n, R. Notario, J. Quijano, C. Snchez, J. Phys. Chem. A 2002, 106,

6618.[47] J. E. Lee, W. Choi, B. J. Mhin, J. Phys. Chem. A 2003, 107, 2693.[48] O. V. Dorofeeva, V. S. Yungman, J. Phys. Chem. A 2003, 107, 2848.[49] J. S. Murray, P. Politzer, J. Mol. Struct. (Theochem) 1988, 163, 111.[50] T. Fujii, K. Tanaka, H. Tokiwa, Y. Soma, J. Phys. Chem. A 1996, 100, 4810.[51] K. Tuppurainen, J. Ruuskanen, Chemosphere 2000, 41, 843.[52] K. Tuppurainen, I. Halmen, J. Ruuskanen, Chemosphere 1996, 32, 1349.[53] Y. Okamoto, M. Tumonari, J. Phys. Chem. A 1999, 103, 7686.[54] K. Tuppurainen, J. Ruuskanen, Chemosphere 1999, 38, 1825.[55] A. A. Farajian, M. Mikami, P. Ordej�n, K. Tanabe, J. Chem. Phys. 2001,

115, 6401.[56] Y. Okamoto, Chem. Phys. Lett. 1999, 310, 355.[57] J. Eun, W. Choi, B. J. Mhin, K. Balasubramanian, J. Phys. Chem. A 2004,

108, 607.[58] Gaussian 98 (Revision A.6) M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E.

Scuseria, M. A. Robb, J. R. Cheeseman, V. G. Zakrzewski, J. A. Montgom-ery, R. E. Stratmann, J. C. Burant, S. Dapprich, J. M. Millam, A. D. Daniels,K. N. Kudin, M. C. Strain, O. Farkas, J. Tomasi, V. Barone, M. Cossi, R.Cammi, B. Mennucci, C. Pomelli, C. Adamo, S. Clifford, J. Ochterski, G. A.Petersson, P. Y. Ayala, Q. Cui, K. Morokuma, D. K. Malick, A. D. Rabuck, K.Raghavachari, J. B. Foresman, J. Cioslowski, J. V. Ortiz, B. B. Stefanov, G.Liu, A. Liashenko, P. Piskorz, I. Komaromi, R. Gomperts, R. L. Martin, D. J.Fox, T. Keith, M. A. Al-Laham, C. Y. Peng, A. Nanayakkara, C. Gonzalez, M.Challacombe, P. M. W. Gill, B. Johnson, W. Chen, M. W. Wong, J. L.Andres, C. Gonzalez, M. Head-Gordon, E. S. Replogle, J. A. Pople, Gaussi-an, Inc. , Pittsburgh, PA, 1998.

[59] A. D. Becke, Phys. Rev. A 1988, 38, 3098.[60] C. Lee, W. Yang, R. G. Parr, Phys. Rev. B 1988, 37, 785.[61] A. D. Becke, J. Chem. Phys. 1993, 98, 5648.[62] W. J. Hehre, L. Radom, J. A. Pople, P. v. R. Schleyer, Ab Initio Molecular Or-

bital Theory, Wiley, New York, 1986.[63] H. B. Schlegel, J Comput. Chem. 1982, 3, 214.[64] C. Gonzalez, H. B. Schlegel, J. Chem. Phys. 1989, 90, 2154.[65] C. Gonzalez, H. B. Schlegel, J. Phys. Chem. 1990, 94, 5523.[66] R. Krishnan, J. S. Binkley, R. Seeger, J. A. Pople, J. Chem. Phys. 1980, 72,

650.[67] A. D. McLean, G. S. Chandler, J. Chem. Phys. 1980, 72, 5639.[68] D. A. McQuarrie, Statistical Mechanics, Harper & Row, New York, 1986.[69] F. P. Boer, M. A. Neuman, F. P. Van Remoortere, P. P. North, H. W. Rinn,

Adv. Chem. Ser. 1973, 120, 14.

Received: June 9, 2006Revised: September 5, 2006


R. L�pez et al.


Documents

Formation of Trichlorinated Dibenzo-p-dioxins from 2,4-Dichlorophenol and 2,4,5-Trichlorophenolate: A Theoretical Study