thermochemistry and photodissociation studies of [col]+ and [col2]+, l=pyrrole, furan, thiophene and...

JOURNAL OF MASS SPECTROMETRY, VOL. 32, 475È482 (1997)

Thermochemistry and Photodissociation Studies ofand Furan,[CoL ]‘ [CoL

2]‘, L = Pyrrole,

Thiophene and Selenophene

Vajira K. Nanayakkara and Ben S. Freiser*H. C. Brown Laboratory of Chemistry, Purdue University, West Lafayette, Indiana 47907, USA

[CoL ]‘ and (L = pyrrole, furan, thiophene and selenophene) were generated in a prototype Nicolet-[CoL2

]‘FTMS 1000 Fourier transform mass spectrometer, where they were irradiated with light from a 2.5 kW Xe arclamp. Photodissociation thresholds, obtained using energy cut-o† Ðlters, yielded the bond energiesD0(Co‘wpyrrole) = 59 » 3 kcal mol—1, D0(Co‘wfuran) = 57 » 3 kcal mol—1, D0(Co‘wthiophene) = 61 » 3kcal mol—1 and D0(Co‘wselenophene) = 64 » 3 kcal mol—1 (1 kcal = 4.184 kJ), as well asD0( [Co(pyrrole) ]‘wpyrrole) O 48 » 3 kcal mol—1, D0( [Co(furan) ]‘wfuran) O 46 » 3 kcal mol—1,D0( [Co(thiophene) ]‘wthiophene) O 49 » 3 kcal mol—1 and D0( [Co(selenophene) ]‘wselenophene) O 51 » 3kcal mol—1. The photoappearance threshold for the CO loss product ion from [Co(furan) ]‘, yielded[CoC

3H

4]‘,

33 » 6 or 32 » 6 kcal mol—1, depending on whether was allene or propyne,D0(Co‘wC3H

4) = either C

3H

4respectively. Where possible, these bond energy measurements were further corroborated by ligand displacementreactions and competitive collision-induced dissociation experiments.

J. Mass Spectrom. 32, 475È482 (1997)No. of Figures : 8 No. of Tables : 1 No. of References : 30

KEYWORDS: Fourier transform mass spectrometry ; photodissociation ; transition metal ions ; bond energies ; aromaticheterocycles

INTRODUCTION

In recent decades there has been growing interest in thechemistry of heterocycles coordinated to transitionmetal ion centers such as Ru(I), Os(II) and Mn(I).1Among these heterocyclic ligands, pyridine has been themost extensively studied in condensed media.2 Pyridine,like benzene, has a high tendency to form stable com-plexes owing to its good n-donor properties.3 Althoughthe Ðve-membered heterocyclic ligands are poorer n-electron donors than benzene and pyridine, many exam-ples can be found in the literature of such complexes inthe condensed phase,2 with some examples reportedfrom gas-phase studies.4 Among the Ðve-membered het-erocycles, pyrroles, furans, thiophenes and selenophenesrepresent important classes of compounds. The pyrrolering system, for example, is widely distributed in natureand found as subunits of heme, chlorophylls, vitamin

bile pigments, porphyrins, antibiotics and variousB12 ,polymeric systems.1c,5

Thiophenes, thiols and cyclic sulÐdes present in fossilfuels are mainly responsible for the poisoning of heter-ogeneous catalysts during the hydrogenation processutilized by the petroleum industry.6 In crude oil reÐn-ing, metal ion-promoted degradation of thiophenes is amajor process in the removal of organosulfur impuritiespresent in fossil fuels.7 Owing to the importance of aro-matic heterocycles, especially pyrrole and thiophene, inindustrial applications, as well as in the biomedical sci-

ences, extensive studies on the reactivity and bindingproperties of these heterocycles with di†erent metalcenters have been carried out in the condensed phase byseveral research groups.1b,8 Recently, Bakhtiar andJacobson studied the gas-phase reactions of Fe` and[FeL]` [(L \ O, C4H6 , c-C5H6 , c-C5H5 , C6H6 ,

with furan, thiophene and pyrrole.4C5H4(xCH2)]As part of their study, bond energies were measuredincluding D0(Fe`wpyrrole) B D0(Fe`wthiophene)B

kcal mol~1 andD0(Fe`wC4H6) B 48 ^ 5D0(Fe`wfuran) [ 39.9^ 1.4 kcal mol~1 (1 kcal\4.184 kJ). In this work, we examined the bindingproperties of [CoL]` and (where L \ pyrrole,[CoL2]`furan, thiophene and selenophene) in the gas phase.

Absolute bond energy measurements are of primaryimportance in understanding the thermodynamic andmechanistic aspects of various processes which mayoccur during a reaction. Appearance potential measure-ments,9 thresholds for endothermic ionÈmolecule reac-tions,10 competitive collision-induced dissociation(CID),11 kinetic energy release distribution measure-ments12 and photodissociation threshold experiments13are some of the gas-phase techniques that have beenwidely used for the determination of metalÈligand bondenergies. Of these techniques, determining thethresholds of endothermic reactions by using guided ionbeams, pioneered by Beauchamp, Armentrout and theirco-workers, has been found to be the most useful todate, providing a large number of high-quality absolutebond energy values.14 Additionally, in the past decade,

CCC 1076È5174/97/050475È08 $17.50 Received 30 September 1996( 1997 by John Wiley & Sons, Ltd. Accepted 25 November 1996

476 V. K. NANAYAKKARA AND B. S. FREISER

theoretical calculations have come a long way towardsproviding accurate bond energies, as well as detaileddescriptions of the bonding.15

In our laboratory, photodissociation has been foundto be an e†ective tool for obtaining absolute bond ener-gies in metal ionÈligand complexes.16 Photodissociationthresholds are measured by monitoring the appearanceof M` in reaction (1) as a function of wavelength.

[ML]`] hl] M`] L (1)

If the binding energies are in a region where there is ahigh density of low-lying excited states, then the ionsabsorb photons at this energy and dissociate to yieldthe metal ion.13,16 In such a situation, the photo-dissociation threshold provides a good estimate of theabsolute thermodynamic bond energy. In contrast, if thelow-lying excited states lie in a range above the bindingenergy, the photodissociation threshold measured isspectroscopic and yields only an upper limit of the bondenergy. The only way to know whether a photo-dissociation threshold is spectroscopic or thermo-dynamic is to compare the value to the bond energyobtained by at least one other method. In this work,ionÈmolecule reaction and CID bracketing experimentswere used, when possible, to provide corroboratinginformation.

It should also be mentioned that contributions fromsequential multiphoton absorption processes17 mayyield erroneously low bond energy values. Based on thegenerally good agreement of the photodissociationvalues with those obtained from the bracketing experi-ments, however, we do not feel that multiphoton disso-ciation occurs in these experiments to any signiÐcantextent.

EXPERIMENTAL

The instrument used in this work was a Nicolet (nowFinnigan FT/MS, Madison, WI, USA) prototypeFTMS-1000 Fourier transform ion cyclotron resonance(FTICR) mass spectrometer, equipped with a 5.2 cmcubic cell which is placed in between the poles of aWalker 15 in electromagnet maintained at D0.9 T. Theprinciples, theory and performance of FTICR massspectrometers have been described previously.18 Thetwo transmitter plates in our instrument are composedof 80% transmittance stainless-steel screens which allowthe ions in the cell to be irradiated with light fromvarious sources. Co` was generated by focusing thefundamental line (1.06 km) of a Quanta Ray Nd :YAGlaser on to a high-purity cobalt target which wasmounted on one of the transmitter plates.19 Prior toreaction with the reagent, the Co` ions generated werethermalized by collisions with argon for 1È2 s. Theargon was introduced into the cell through a leak valveat a background pressure of 3.5 ] 10~5 Torr (1Torr \ 133.3 Pa), as measured by an uncalibratedBayardÈAlpert ionization gauge. The [CoL]` and

were also cooled by collisions with argon and[CoL2]`the reagent gas during the 5 s irradiation time. All of the

reagents were introduced into the cell through aGeneral Valve Series 9 pulsed solenoid valve.20 Thepulsed reagent gas Ðlled the vacuum chamber to amaximum pressure of about 4 ] 10~5 Torr with abouta 150 ms rise time and was pumped away by a high-speed 6-in di†usion pump in about 300 ms. The[CoL]` (L\ pyrrole, furan, thiophene andselenophene) complex thus formed was then isolated bya swept double-resonance pulse.21 In order to make the

complex, the pulsed valve opening time was[CoL2]`increased from 2 to 20 ms, yielding a maximumchamber pressure of D8 ] 10~5 Torr.

An Osram Sylvania XBO model 2.5 kW Xe arc lampwas used in conjunction with di†erent energy cut-o†Ðlters to determine thresholds.22 A Spectra-PhysicsModel 2030-18 argon ion laser was also used in onecase for determining D0(Co`wpyrrole), as a furthercheck of the arc lamp results. This ion was chosen sinceits threshold fell conveniently in the wavelength regionof the laser. The beam was focused into the cell througha sapphire window with three optical mirrors. The ionswere irradiated for 5 s in all of the experiments and thephotoproduct peak intensities were blank subtracted,normalized and plotted against the cut-o† energies todetermine the thresholds. With the argon ion laser, indi-vidual lines in the wavelength range 458È529 nm wereselected and operated at 1 W maximum output power,as measured using a Spectra-Physics Model 2030-18power meter. All chemicals were obtained from com-mercial sources and used without further puriÐcationexcept for multiple freeze-pumpÈthaw cycles to removenon-condensible gases.

RESULTS AND DISCUSSION

(x = 1, 2)[Co(pyrrole)x

]‘

Photodissociation of [Co(pyrrole)]` proceeds exclu-sively by reaction (1) to generate Co`. The photoap-pearance spectrum for Co` was obtained by irradiating[Co(pyrrole)]` with light from the arc lamp. A sharpincrease in metal ion abundance was observed in theregion from 57È60 kcal mol~1 [Fig. 1(a)]. Photo-dissociation of [Co(pyrrole)]` with the individual linesof the argon ion laser is shown in Fig. 1(b) for compari-son. The Co` abundance remains zero or near zero at502 nm and, thereafter, a sharp increase in metal ionabundance is observed using the shorter wavelengthlines, from which D0(Co`wpyrrole) \ 59 ^ 3 kcalmol~1 is assigned. Hence there is excellent agreementbetween the arc lamp and laser results. In addition, thephotodissociation results are further corroborated byusing ligand displacement reactions and competitiveCID experiments, as discussed below.

[Co(pyrrole)]` reacts readily with benzene byligand displacement to form [Co(benzene)]`. Thereverse reaction of [Co(benzene)]` with pyrrole,however, yields only the condensation product,

( 1997 by John Wiley & Sons, Ltd. JOURNAL OF MASS SPECTROMETRY, VOL. 32, 475È482 (1997)

PHOTODISSOCIATION STUDIES OF [Co(C4H4X)1, 2]` , X\ NH, O, S, SE 477

(a)

(b)

Figure 1. Photoappearance spectrum of Co½ from photo-dissociation of ÍCo(pyrrole)Ë½ using (a) the arc lamp and cut-offfilters and (b) individual lines of the argon ion laser. The data arenormalized per photon at each wavelength.

[Co(benzene)(pyrrole)]`. Collision-induced dissociationof the condensation product yields [Co(benzene)]`,exclusively, by loss of pyrrole over the entireenergy range studied. These results indicate

that D0(Co`wpyrrole) \ D0(Co`wbenzene)\ 61.1^ 2.5 kcal mol~1.23 A lower limit forD0(Co`wpyrrole) is obtained from the reaction of

with pyrrole. generated from theCoNH3` CoNH3`,reaction of Co` with pulsed-in reacts readily withNH3 ,pyrrole to yield mainly the ligand displacement product,[Co(pyrrole)]`, and some of the condensation product.Further, the reverse reaction of [Co(pyrrole)]` with

yields some ligand displacement, but condensationNH3is the major reaction channel observed. Collision-induced dissociation of at low[Co(NH3)(pyrrole)]`laboratory kinetic energies (\26 eV), yields only NH3loss. As the laboratory kinetic energy is increased, lossof pyrrole is also observed (Fig. 2). These results indi-cate that D0(Co`wpyrrole) PD0(Co`wNH3) \ 58.8kcal mol~1.24 All of the above results are consistentwith the photodissociation threshold value forD0(Co`wpyrrole) \ 59 ^ 3 kcal mol~1.

At longer pulsed-valve opening time (30 ms), the sec-ondary product ion is observed.[Co(pyrrole)2]`was isolated, using frequency sweep[Co(pyrrole)2]`double-resonance techniques,21 and irradiated withthe light from the arc lamp at di†erent energy cut-o†s. The sole photoproduct, [Co(pyrrole)]`, beginsto appear using the 48 kcal mol~1 energy cut-o† Ðlter(Fig. 3). The relative abundance of [Co(pyrrole)]`gradually increases from 48È84 kcal mol~1, atwhich point a sharp increase is observed. Fromthe above photodissociation data, we assignD0([Co(pyrrole)]`wpyrrole) \ 48 ^ 3 kcal mol~1.This result is further supplemented by liganddisplacement reactions. reacts with[Co(pyrrole)2]`

Figure 2. Variation of relative abundance as a function of energy(center of mass frame) for CID of atÍCo(NH

3)(pyrrole)Ë½

3 Ã10É5 Torr argon target gas.



Figure 3. Photoappearance spectrum of ÍCo(pyrrole)Ë½ fromphotodissociation of using the arc lamp andÍCo(pyrrole)

2Ë½

cut-off filters.

by sequential ligand displacement toNH3yield and[Co(pyrrole)(NH3)]` Co(NH3)2`,indicating that D0([Co(pyrrole)]`wpyrrole)\

andD0([Co(pyrrole)]`wNH3) D0(Co(NH3)`wpyrrole)No reaction is observed for\ D0(Co(NH3)`wNH3).the reverse process of with pyrrole. TheseCo(NH3)2`results suggest that D0([Co(pyrrole)]`wpyrrole)

kcal mol~1.23 Un-\ D0(Co(NH3)`wNH3) \ 61 ^ 2fortunately, a lower limit could not be assigned due tothe absence of reference bond energy data. Thus, whileD0([Co(pyrrole)]`wpyrrole) O 48 ^ 3 kcal mol~1 isprobably a good estimate of the absolute bond energy,it represents strictly only an upper limit.

(x = 1, 2)[Co(furan)x]‘

In contrast to pyrrole, where monomer and dimer for-mation are the only reaction channels observed, withfuran, in addition to these channels, CO loss (maximumof 40% under the reaction conditions studied) is alsoobserved. Interestingly, the analogous reaction is notobserved with Fe`, but is with various [FeL]` species

and The structure(L\C4H6, c-C5H6 C5H4(xCH2)].4of the CO loss product ion, was probed by[CoC3H4]`,CID and other ionÈmolecule reactions, some of whichare described here. Complete ligand displacement isobserved when reacts with propene, sug-[CoC3H4]`

gesting that the moiety bound to Co` is intactC3H4and that D0(Co`wC3H4) \ D0(Co`wpropene)\ 48kcal mol~1.12a Collision-induced dissociation of^ 2

yields Co` as the sole product at all of the[CoC3H4]`energies selected. These results are consistent with a[Co(allene)]` and/or [Co(propyne)]` structure for

To date we have been unable to distinguishCoC3H4]`.between these two possibilities.

Photodissociation of [Co(furan)]` yields two photo-products, Co` and the CO loss product, [CoC3H4]`.Co` begins to appear at 56 kcal mol~1 yieldingD0(Co`wfuran) \ 57 ^ 3 kcal mol~1 (Fig. 4).[Co(propene)]`, generated from the reaction ofCo` with pulsed-in propene, reacts with furan toyield the ligand displacement product, [Co(furan)]`.The reverse reaction of [Co(furan)]` with propeneyields only the condensation product. Upon CID,[Co(furan)(propene)]` yields [Co(furan)]`, exclu-sively, over the range of laboratory kinetic energiesstudied. These results yield a lower limit ofD0(Co`wfuran) [ D0(Co`wpropene)\ 48 ^ 2 kcalmol~1.12a [Co(furan)]` reacts with by ligand dis-NH3placement to form In the reverse reaction ofCo(NH3)`.

with furan, condensation is observed, exclu-Co(NH3)`sively. CID on yields the furan[Co(furan)(NH3)]`loss product, These results yield an upperCo(NH3)`.limit of kcalD0(Co`wfuran) \D0(Co`wNH3) \ 58.8mol~1.24 Hence, D0(Co`wfuran) \ 57 ^ 3 kcal mol~1is in accordance with the above ionÈmolecule reactionsand CID experiments.

The CO loss photoproduct has a photoappearancethreshold at 52 ^ 3 kcal mol~1 [Fig. 4(b)]. This

Figure 4. Photoappearance spectrum of (a) Co½ and (b)from photodissociation of ÍCo(furan)Ë½, using cut-offÍCoC

3H

4Ë½

filters.



threshold can be used to calculate D0(Co`wC3H4).First, kcal mol~1 is*Hf([Co(furan)]`)\ 218 ^ 3obtained from Eqn (2), using kcal*Hf(Co`) \ 283mol~1,25a kcal mol~125b and*Hf(furan) \ [8.3^ 0.1D0(Co`wfuran) \ 57 ^ 3 kcal mol~1 obtained inthis work. Next, kcal*Hf([CoC3H4]`)\ 296 ^ 6mol~1 is calculated from Eqn (3), using *Hf(CO)\[26.42 kcal mol~125a and 52 ^ 3 kcal mol~1 fromthe photoappearance threshold for CO loss from[Co(furan)]`. Finally, Eqn (4) yields D0(Co`wallene)\33 ^ 6 kcal mol~1 and/or D0(Co`wpropyne) \ 32 ^ 6kcal mol~1 based on either *Hf(allene)26 \ 45.6 ^ 0.2kcal mol~1 or kcal*Hf(propyne)25b \ 44.6 ^ 0.5mol~1. The values reported above are consistentwith the report by Haynes and Armentroutthat D0(Co`wallene)[ 18.7^ 2.1 kcal mol~1 and withour observation above of D0(Co`wallene)\D0(Co`wpropene)\ 48 ^ 2 kcal mol~1.27

*Hf([Co(furan)]`)\ *Hf(Co`) ] *Hf(furan)

[ D0(Co`wfuran) (2)

*Hf([CoC3H4]`) \ *Hf([Co(furan)]`) [ *Hf(CO)

] 52 ^ 3 kcal mol~1 (3)

D0(Co`wC3H4) \ *Hf([Co(C3H4)]`)[ *Hf(Co`)

[ *Hf(C3H4) (4)

The thermochemistry for CX loss (X\ NH, O and S)in reaction (5) of Co` with pyrrole, furan and thiopheneis determined by Eqn (6) and listed in Table 1. All of theheats of formation required to solve Eqn (6) are avail-able except for which we deter-*Hf([Co(C3H4)]`),mined above to be 296 ^ 6 kcal mol~1. As can be seenin Table 1, reaction (5) is only exothermic for furan,which is consistent with the observed reactions of Co`with the four heterocyclic compounds. Although

is not available for the determination of*Hf(CSe) *Hrxnfor CSe loss, based on the thermochemistry for CS loss,CSe loss is presumably also endothermic. In any event,CSe loss is not observed in the reaction of Co` withselenophene.

Co`] C4H4X] [Co(C3H4)]`] CX (5)

where X \ NH, O, S and Se

*Hrxn\ *Hf([Co(C3H4)]`) ] *Hf(CX)[ *Hf(Co`)

[ *Hf(C4H4X) (6)

Table 1. Reaction enthalpy for CX loss (X = NH, O(DHrxn

)and S) in reactions of Co‘ with aromatic hetero-cycles : Co‘ + C

4H

4X Ç [Co(C

3H

4) ]‘ + CXa

DHf(CX) DH

f(C

4H

4X) DH

rxnC

4H

4X X (kcal molÉ1) (kcal molÉ1) (kcal molÉ1)

Pyrrole NH ½32.325a 25.9À0.126 ½19.4 À6.1

Furan O É26.425a É8.3À0.125b É5.1 À6.1

Thiophene S ½6426 ½27.5 À0.126 ½49.5 À6.1

kcal molÉ1 as calculated in thisa DHf(ÍCo(C

3H

4Ë½) ¼296 À6

work. kcal molÉ1.25aDHf(Co½) ¼283

Figure 5. Photoappearance spectrum of ÍCo(furan)Ë½ fromphotodissociation of using cut-off filters.ÍCo(furan)

2Ë½

Photodissociation of yields[Co(furan)2]`[Co(furan)]`, exclusively, as the photoproduct. Theonset for the photoappearance of [Co(furan)]` is46 ^ 3 kcal mol~1 (Fig. 5). Owing to a lack of bondenergy data for disubstituted Co`, the value obtainedhere for D0([Co(furan)]`wfuran) was not examined byother ionÈmolecule reactions and CID experiments.Thus, we assign D0([Co(furan)]`wfuran) O 46 ^ 3 kcalmol~1.

(x = 1, 2)[Co(thiophene)x

]‘

Based on the photoappearance of Co` [Fig. 6(a)],D0(Co`wthiophene) is measured as 61^ 3 kcal mol~1.[Co(thiophene)]` is observed to react with toNH3yield the condensation product predominantly, with

formation reaching a maximum of D47%Co(NH3)2`under the reaction conditions studied. CID ofresulted in the formation of[Co(thiophene)(NH3)]`[Co(thiophene)]` from loss of gener-NH3. CoNH3`,

ated from the reaction of Co` with pulsed-in NH3 ,however, reacts with thiophene to yield the ligand dis-placement product, [Co(thiophene)]`. At sufficientlylonger reaction times ([400 ms), some condensationis also observed. These results indicate that

kcalD0(Co`wthiophene)[D0(Co`wNH3) \ 58.8mol~1.24 When [Co(benzene)]` is allowed to react withthiophene, condensation is the major reaction channel



Figure 6. Photoappearance spectrum of (a) Co½ from photo-dissociation of ÍCo(thiophene)Ë½ and (b) ÍCo(thiophene)Ë½ fromphotodissociation of using cut-off filters.ÍCo(thiophene)

2Ë½,

observed. Upon CID, the condensation product ionyields [Co(benzene)]` from loss of thiophene. On theother hand, [Co(thiophene)]` reacts with benzene toyield the ligand displacement product, [Co(benzene)]`.Based on these results, D0(Co`wthiophene)\D0(Co`wbenzene) \ 61.1^ 2.5 kcal mol~1.23a Thus,D0(Co`wthiophene)\ 61 ^ 3 kcal mol~1 measuredfrom arc lamp photodissociation is a reasonable valuefor the bond energy.

The photoappearance spectrum obtained by moni-toring [Co(thiophene)]`, upon irradiation of

is shown in Fig. 6(b). A sharp[Co(thiophene)2]`,threshold for the appearance of [Co(thiophene)]`begins at 49^ 3 kcal mol~1. Similarly to

and only an upper[Co(pyrrole)2]` [Co(furan)2]`,limit, D0([Co(thiophene)]`wthiophene)O 49 ^ 3 kcalmol~1, is assigned.

(x = 1, 2)[Co(selenophene)x

]‘

The photoappearance data of Co` from[Co(selenophene)]` are shown in Fig. 7(a). From thephotodissociation threshold, D0(Co`wselenophene)\64 ^ 3 kcal mol~1 is assigned. When reactsCo(NH3)`with selenophene, complete ligand displacement isobserved. In contrast, [Co(selenophene)]` yields onlythe condensation product with CIDNH3 .

Figure 7. Photoappearance spectrum of (a) Co½ from the photo-dissociation of ÍCo(selenophene)Ë½ and (b) ÍCo(selenophene)Ë½

from photodissociation of using cut-offÍCo(selenophene)2Ë½,

filters.

of yields predominantly[Co(selenophene)(NH3)]`[Co(selenophene)]` from loss of TheseNH3 .results indicate that D0(Co`wselenophene)[

kcal mol~1.24 [Co(benzene)]`D0(Co`wNH3) \ 58.8reacts with selenophene to yield a condensationproduct which, upon CID, produces [Co(benzene)]`.Selenophene ligand is displaced in the reverse reactionof [Co(selenophene)]` with benzene. These resultsare consistent with D0(Co`wselenophene)\D0(Co`wbenzene) \ 61.1^ 2.5 kcal mol~1.22a Hencethe photodissociation threshold is at least nearlythermodynamic and D0(Co`wselenophene)\ 64 ^ 3kcal mol~1, measured from arc lamp photodissociation,is a good estimate of the absolute bond energy.

shows a sharp photo-[Co(selenophene)2]`dissociation threshold starting at 51 ^ 3 kcal mol~1[Fig. 7(b)]. As discussed for the other [CoL2]`complexes (L\ pyrrole, furan and thiophene),D0([Co(selenophene)]`wselenophene)O 51 ^ 3 kcalmol~1 is, in the absence of other corroborating experi-ments, strictly only an upper limit for the bond energyof [Co(selenophene)]` to selenophene.

Competitive collision-induced dissociation on Co‘-boundheterodimers

Based on the photodissociation threshold measure-ments, the bond energies of [CoL]` vary systematicallyas follows : D0(Co`wfuran)\D0(Co`wpyrrole) \D0(Co`wthiophene)\ D0(Co`wselenophene). Giventhe error bars and the closeness of the values, however,



additional support for this relative bond energy orderwas obtained by competitive ligand displacement reac-tions and CID.

First, [Co(pyrrole)]` was allowed to react with furan,yielding only the condensation product. Upon CID,[Co(pyrrole)(furan)]` yields mainly [Co(pyrrole)]`together with some [Co(furan)]` [Fig. 8(a)]. In thereverse process, [Co(furan)]` reacts with pyrrole andyields only the ligand displacement product,[Co(pyrrole)]`. These results suggest that D0(Co`wfuran) \ D0(Co`wpyrrole), in agreement with thephotodissociation threshold measurements.

The reactions of [Co(pyrrole)]` with thiophene and[Co(thiophene)]` with pyrrole were examined next.[Co(pyrrole)]` reacts with thiophene to generatepredominantly the ligand displacement product,[Co(thiophene)]`. At longer reaction times ([400 ms),the condensation product ion (D18% maximum underthe reaction conditions studied) is also observed.The reaction of [Co(thiophene)]` with pyrroleyields [Co(thiophene)(pyrrole)]` as the major pro-duct and [Co(pyrrole)]` as the minor product ion atall of the reaction times studied. CID on thecondensation product, [Co(pyrrole)(thiophene)]`, yields[Co(thiophene)]` as the major product with some[Co(pyrrole)]` formation [Fig. 8(b)]. Similarly,[Co(selenophene)]` reacts with thiophene to yield thecondensation product, [Co(selenophene)(thiophene)]`.

Figure 8. Variation of relative abundance as a function of energy(center of mass frame) for CID of Co½-bound heterodimers (a)ÍCo(pyrrole)(furan)Ë½, (b) ÍCo(pyrrole)(thiophene)Ë½ and (c)ÍCo(thiophene)(selenophene)Ë½ at 3.5 Ã10É5 Torr argon targetgas.

CID of [Co(thiophene)(selenophene)]` yields predomi-nantly [Co(selenophene)]`, with [Co(thiophene)]`observed as a minor product ion above D3 eV center ofmass energy [Fig. 8(c)]. In accordance with the photo-dissociation threshold measurements, these resultsindicate that D0(Co`wfuran) \ D0(Co`wpyrrole) \D0(Co`wthiophene) \ D0(Co`wselenophene).

CONCLUSION

The bond energies obtained from photodissociation,D0(Co`wpyrrole) \ 59 ^ 3 kcal mol~1, D0(Co`wfuran) \ 57 ^ 3 kcal mol~1, D0(Co`wthiophene)\61 ^ 3 kcal mol~1 and D0(Co`wselenophene)\64 ^ 3 kcal mol~1, are in accordance with the upperand lower limits assigned by ligand displacement reac-tions and CID experiments and, therefore, representthermodynamic thresholds. Previous studies carried outin condensed phases by several research groups haveshown that pyrrole n-bonds to metal centers.1 Withthiophene, however, several n-bonding modes, as wellas p-bonding via the S atom, have been suggested.28The experimental data and bond energies obtained inthis study do not allow us to distinguish the modeof bonding present in these [ML]` complexes. Thesimilarities of the bond energies, especially topyrrole, however, prompt us to suggest that the mostlikely bonding mode is n-bond formation instead ofheteroatom bound p-bond formation. These resultsare also consistent with those obtained byBakhtiar and Jacobson4 indicating thatD0(Fe`wpyrrole) B D0(Fe`wthiophene)B 48 ^ 5 kcalmol~1. Considering that D0(Co`wbenzene)23a [ D0(Fe`wbenzene)23cB 11 kcal mol~1,the higher values we obtained for the cobalt complexesare certainly reasonable.

For the species, photodissociation yielded[CoL2]`the bond energy limits D0([Co(pyrrole)]`wpyrrole)O48 ^ 3 kcal mol~1, D0([Co(furan)]`wfuran) O 46 ^ 3kcal mol~1, D0([Co(thiophene)]`wthiophene)O49 ^ 3 kcal mol~1 and D0([Co(selenophene)]`wselenophene)O 51 ^ 3 kcal mol~1. The observationthat the second ligand is more weakly bound to the Cometal center in these species is in accordance with pre-vious experiments carried out on systems such as

(M\ Fe, Co, Ni and Cu) where x \ 1,[M(C6H6)x]`2.27 In order to satisfy the 18-electron rule, the forma-tion of stable [Co(pyrrole)2]`, [Co(furan)2]`,

and complexes[Co(thiophene)2]` [Co(selenophene)2]`suggests that, assuming the bonding mode of the Ðrstligand is g6, then that of the second ligand is gx (x O 5)or heteroatom bound and, thus, more weakly bonded.This is reasonable considering that a variety of di†erentbinding modes for these heterocycles have been charac-terized.29 There are also other reasons for reducedD0([ML]`wL) in species, including steric and[ML2]`electronic factors. IonÈdipole and ionÈinduced dipoleinteractions decrease as the charge is delocalized to anyextent on the ligand in the monomer, [ML]`.15a,30Second, the steric e†ects caused by the ligand alreadypresent increases the ligandÈligand repulsions. Hence,



the reduced electrostatic interactions and crowdinge†ects are two other important factors which may causeD0([ML]`wL)\ D0(M`wL).

Acknowledgements

Acknowledgement is made to the Division of Chemical Sciences in theOffice of Basic Energy Sciences in the United States Department of

Energy(under grant DE-FG02-87ER13766) for supporting thisresearch.

REFERENCES

1. (a) K. M. K. Dailey, T. B. Rauchfuss, A. L. Rheingold andG. P. A. Yap, J. Am. Chem. Soc. 117, 6396 (1995); (b) L. M.Hodges, J. Gonzalez, J. I. Koontz, W. H. Myers and W. D.Harman, J. Org. Chem. 58, 4788 (1993); (c) R. Cordone,W. D. Harman and H. Taube, J. Am. Chem. Soc. 111, 5969(1989); (d) B. Chaudret and F. A. Jalon, J. Chem. Soc.,Chem.Commun. 711 (1988).

2. P. George, Chem. Rev. 75, 85 (1975), and references citedtherein.

3. (a) M. J. Cook, A. R. Katritzky and P. Linda, Adv. Heterocycl .Chem. 17, 255 (1974); (b) B. J. Deelman, W. M. Stevels,J. H. Teuben, M. T. Lakin and A. L. Spek, Organometallics 13,3881 (1994).

4. R. Bakhtiar and D. B. Jacobson, J. Am. Soc. Mass Spectrom.7, 938. (1996).

5. (a) A. Jones, Chemistry of Heterocyclic Compounds, Vol. 48.Wiley, New York (1990); (b) W. H. Myers, J. I. Koontz andW. D. Harman, J. Am. Chem. Soc. 114, 5684 (1992); (c)W. T. Brockmann and J. M. Tour, J . Am. Chem. Soc. 117,4437 (1995).

6. T. Inul and Y. Tanabe, J.Catal . 52, 375 (1978).7. T. B. Rauchfuss, Prog. Inorg.Chem. 39, 259 (1991).8. (a) E. W. Abel and C. J. Towers, J . Chem. Soc., Dalton Trans.

814 (1979); (b) C. Segard, C. Pommier, B. P. Roques and G.Guiochon, J. Organomet . Chem. 77, 49 (1974); (c) R. B.King and J. Efraty, J .Organomet . Chem. 20, 264. (1969).

9. E. Murad, J.Chem.Phys. 73, 1381 (1980).10. (a) P. B. Armentrout and J. L. Beauchamp, J. Am. Chem. Soc.

103, 784 (1981); (b) P. B. Armentrout, L. F. Halle and J. L.Beauchamp, J.Am.Chem.Soc. 103, 6501 (1981).

11. S. A. McLuckey, A. E. Schoen and R. G. Cooks, J. Am. Chem.Soc. 104, 848 (1982).

12. (a) M. A. Hanratty, J. L. Beauchamp, A. J. Illies, P. vanKoppen and M. T. Bowers, J . Am. Chem. Soc. 110, 1 (1988);(b) P. A. M. van Koppen, D. B. Jacobson, A. J. Illies, M. T.Bowers, M. A. Hanratty and J. L. Beauchamp, J. Am. Chem.Soc. 111, 1991 (1989).

13. (a) S. W. Buckner and B. S. Freiser, Polyhedron 7, 1583(1988); (b) W. J. van der Hart, Mass Spectrom. Rev. 8, 237(1989).

14. (a) P. B. Armentrout, R. V. Hodges and J. L. Beauchamp,Chem. Phys. 66, 4683 (1977); (b) P. B. Armentrout and R.Georgiadis, Polyhedron 7, 1573 (1988); (c) P. B. Armentrout,in Gas Phase Inorganic Chemistry , edited by D. H. Russell,p. 1. Plenum Press, New York (1989); (d) P. B. Armentroutand B. Kickel, in Organometallic Ion Chemistry , edited by B. S.Freiser, p. 1, Kluwer, Dordrecht (1995); (e) E. R. Fisher, B. L.Kickel and P. B. Armentrout, J . Phys. Chem. 97, 10204(1993).

15. (a) C. W. Bauschlicher, Jr, S. R. Langhoff and H. Partridge, inOrganometallic Ion Chemistry , edited by B. S. Freiser, Kluwer,Dordrecht (1995); (b) G. B. McGaughey, E. L. Stewart andJ. P. Bowen, J. Comput . Chem. 17, 1395 (1996); (c) F. Chanand J. Baker, J . Comput . Chem. 17, 888 (1996); (d) M. W.Schmidt, T. L. Windus and M. S. Gordon, J. Am. Chem. Soc.117, 7480 (1995); (e) C. W. Bauchlicher, Jr, H. Partridge andS. R. Langhoff, J . Phys.Chem. 96, 3273 (1992).

16. (a) S. Afzaal and B. S. Freiser, Chem. Phys. Lett . 218, 254(1994); (b) B. S. Freiser and J. L. Beauchamp, J. Am. Chem.Soc. 98, 3136 (1976); (c) L. M. Roth, B. S. Freiser, C. W.Bauschlicher, Jr, H. Partridge and S. R. Langhoff, J . Am.Chem. Soc. 113, 3274 (1991); (d) Y. A. Ranasinghe, I. B.Surjasasmita and B. S. Freiser, in Organometallic Ion Chem-istry , edited by B. S. Freiser, p. 229, Kluwer, Dordrecht(1995).

17. (a) L. R. Thorne and J. L. Beauchamp, in Gas Phase IonChemistry , edited by M. I. Bowers, Vol. 3, p. 42. AcademicPress, New York (1984); (b) R. C. Dunbar, in Gas Phase IonChemistry , edited by M. I. Bowers, Vol. 3, p. 130. AcademicPress, New York (1984).

18. (a) P. B. Grosshans and A. G. Marshall, Int . J . MassSpectrom. Ion Process . 100, 347 (1990); (b) A. G. Marshall,Acc. Chem. Res. 18, 316 (1985); (c) A. G. Marshall and P. B.Grosshans,Anal . Chem. 63, 215A (1991).

19. (a) R. B. Cody, R. C. Burnier, W. D. Reents, Jr, T. J. Carlin,D. A. McCrery, R. K. Lengel and B. S. Freiser, Int . J . MassSpectrom. Ion Phys. 33, 37 (1980); (b) B. S. Freiser, Talanta32, 697 (1985).

20. T. J. Carlin and B. S. Freiser, Anal . Chem. 55, 571 (1983).21. (a) A. G. Marshall and D. C. Roe, J. Chem. Phys. 73, 1581

(1980); (b) A. G. Marshall, T. C. L. Wang and T. L. Ricca,Chem.Phys. Lett . 108, 63 (1984).

22. (a) D. R. A. Ranatunga and B. S. Freiser, Chem. Phys. Lett .233, 319 (1995); (b) C. J. Cassady, S. Afzaal and B. S.Freiser, Org. Mass Spectrom. 29, 30 (1994); (c) Y. A. Rana-singhe and B. S. Freiser, Chem.Phys. Lett . 200, 135 (1992).

23. (a) F. Meyer, F. A. Khan and P. B. Armentrout, J. Am. Chem.Soc. 117, 9740 (1995); (b) C. W. Bauschlicher, H. Partridgeand R. Langhoff, J . Chem. Phys. 96, 3273 (1992); (c) R. L.Hettich, T. C. Jackson, E. M. Stanko and B. S. Freiser, J . Am.Chem.Soc. 108, 5086 (1986).

24. R. R. Squires, J . Am.Chem.Soc. 111, 4101 (1989).25. (a) D. D. Wagman, W. H. Evans, V. B. Parker, R. H. Schumm,

I. Halow, S. M. Bailey, K. L. Churney and R. L. Nuttall, J .Phys. Chem. Ref . Data 11, Suppl . 1 (1982); (b) J. B. Pedleyand J. Rylance, Sussex–NPL Computer Analysed Thermoche-mical Data : Organic and Organometallic Compounds . Uni-versity of Sussex, Brighton (1977).

26. S. G. Lias, J. E. Bartmess, J. F. Liebman, J. L. Holmes, R. D.Levin and W. G. Mallard, J. Phys. Chem. Ref . Data , 17, Suppl.1, 1988.

27. C. L. Haynes and P. B. Armentrout, Organometallics 13, 3480(1994).

28. (a) R. J. Angelici, Coord. Chem. Rev. 105, 61 (1990), andreferences cited therein ; (b) Y. S. Nekrasov and N. I. Vasyu-kova, J.Organomet . Chem. 122, 227 (1976).

29. (a) A. E. Ogilvy, A. E. Skaugset and T. B. Rauchfuss,Organometallics 8, 2739 (1989); (b) A. E. Skaugset, T. B.Rauchfuss and C. L. Stern, J. Am. Chem. Soc. 112, 2432(1990); (c) J. Chen and R. J. Angleici, Organometallics 9,849 (1990).

30. M. L. Mandich, L. F. Halle and J. L. Beauchamp, J. Am.Chem.Soc. 106, 4403 (1984).


thermochemistry and photodissociation studies of [col]+ and [col2]+, l=pyrrole, furan, thiophene and...

Documents