Hatano Y., Katsumura Y., Mozumder A. (Eds.) Charged Particle and Photon Interactions with Matter - Recent Advances, Applications, and Interfaces

Подождите немного. Документ загружается.

250 Charged Particle and Photon Interactions with Matter

In pulse radiolysis experiments, the benzylsilane radical cations (λ ≈ 310, 500…650nm) and the

benzyl radicals (λ ≈ 320…350nm) are visible, but the trimethylsilyl cations are out of the observa-

tion range (Brede etal., 2004). This behavior has been observed for a variety of differently sub-

stituted silanes, and thus can be generalized (Brede and Naumov, 2006). Corresponding data are

given in Table10.3. In the case of cyclic substituents (uorenyl and xanthenyl trimethylsilanes)

(Karakostas et al., 2005), a convincing relationship with the exibility of the silane substituent can

be established, which, in turn, may allow a calibration of the ratio between rapidly formed radicals

and metastable radical cations. This correlation is in fact one of the main arguments supporting

the hypothesis that the products of the FET are a direct reection of the molecular dynamics. The

examples involving benzyltrimethylsilane-type compounds clearly manifest the generality of the

FET concept. Concerning the rapid dissociation of the primary radical cation, this means that not

only protons, as in phenol, but also any other suitable leaving group may be cleaved instantaneously

if the radical cation prevails in the right stereo-electronic state for this process. In the case of the

above-mentioned

silane compounds, the leaving entity is the trimethylsilyl cation.

table 10.3

Fet

with h

ighly

ubstituted

romatic

Compounds, p

roduct

atios

ccording

to

reaction

(10.9) and dynamics data*

donor

rot,def

[10

−15

valence

[10

−15

(s), singlet,

(Cation),

kJ mol

−1

product ratio

(•+)/(•) remarks

-Ar-CR

-SiMe

1330–2380 37 14–24 40:60 R

1,2

= H, Me, Ph;

SiMe

(57–87) R

= H, PHCO-;

compounds

Si(R)

600 40 — 25:75 Xanthenyl

= Me, Ph;

6 examples

Si(R)

— 40 — 100:0 Fluorenyl

= Me, Ph;

2 examples

(Ph

•

*Endothermic

reaction product

C O

440 33

O–CPh

5.9

(29.3)

90: 10

1393 40

S–CPh

4.2

(43.9)

65:35 Also Ph-S-CPh

* Calculated t-times of one cycle of torsion (wagging) and valence (stretching) (ν

valence

) oscillations (B3LYP/6-31G(d),

scale factor ( f = 0.96); E

(S) denotes the energy barrier of rotation along the Ar–CY axis, and bending motions forthe

singlet

ground state, as well as for the radical cation (in brackets).

Reactivity of Radical Cations in Nonpolar Condensed Matter 251

10.4.4.2 trityl-hetero-substituted aromatics

So far, the FET examples described were those in which the dissociation pattern of the unstable donor

radical cations was uniform and in which only one of the two formed species could be observed,

whereas other species, such as protons or silyl cations, were invisible. Using a massive substitution

at the hetero atom, a complete identication of the dissociation products could be demonstrated.

Hence, we took into account the Ar-Y-CPh

structures with the hetero atoms Y=O, S, N substi-

tuted with the trityl rest. The effect of the deformation motion of such molecules is quite similar to

that shown for phenol (cf. Figure 10.7) (Baidak etal., 2008b). In the planar resonant structure, the

π-electrons are distributed over the whole Ar-Y structure, whereas in the twisted state, the electrons

are more localized at the hetero atom. If, in the ground state, the electron density in the twisted

conformation is localized on the hetero (sulfur) atom, then the structure is extremely unstable in the

ionized

form. In contrast, the planar state leads to metastable radical cations.

the FET experiments with the trityl-substituted compounds, partially unexpected results

were obtained. In the case of the example of the highly substituted 2-naphthyltrimethylsulde, we

obtained

the complete transient spectrum shown in Figure 10.10.

The

metastable (but, nevertheless, relatively short-lived) 2-Np-S-

CPh

absorbs around λ=650nm.

It decays with τ

1/2

≈ 200ns. Furthermore, naphthylthiyl radicals (500nm) and the well-known trityl

species (Faria and Steenken, 1990)

CPh

(410nm) and

•

CPh

(330nm) are observed. The trityl spe-

cies can be easily identied because of their high extinction coefcients and the marked peaks. The

appearance of the trityl cation is expected because of the exothermicity of the dissociation of the C–S

bond, whereas the observation of the trityl radical is astonishing because of the highly endothermic

dissociation energy (see reactions 10.13a and 10.13b). Table 10.3 sums up the relevant data.

→ −

−

2-NpS + CPh = 10 kJ mol , 65% yield

∆H

(10.13a)

2-Np S-CPh

2-NpS + CPh 176 kJ mol , 35% yield

− →

→ =

−

∆H

(10.13b)

An analogous effect has been found for trityl-phenylsulde (Karakostas etal., 2007) and also for

trityl-phenyloxide (Baidak etal., 2008b) (see Table 10.3). For the example of the sulde Np-S-CPh

the FET reaction scheme should be adapted (reaction 10.14). In this reaction sequence, it is assumed

90°

Stable structure

0°

Transition state

–e

–

60°

30°

–e

–

(a)

(b)

Figure 10.9 (a) Transformation of HOMO of the Ar-CH

-Si(Me)

singlet ground state (isocontour = 0.06)

dependent on C-SiMe

-group rotation and (b) atomic spin density distribution of the Ar-CH

-SiMe

radical

cation

(isospin = 0.01) as calculated with B3LYP/6-31G(d).

252 Charged Particle and Photon Interactions with Matter

that the unusual dissociation products are formed by the prompt dissociation of the twisted radical

cation. The

observed species are given in boldface:

[Np-S-CPh ] Np-S-CPh

3 planar 3

metastable

delayed

→ →→→ +

i i

NpS CPhh



Np-S- -CPh + RX RX

(10.14a)

[Np-S-CPh ] [Np-S-CPh ]

3 twisted 3

dissociative immediate

→ →

NpS

lly

, (NpS CPh )+ +

i i

CPh

(10.14b)

The assignment of the endergonic products to the dissociation of the twisted radical cation needs an

energetic justication, which is given with the help of a potential energy scheme (see Figure 10.11).

The low-lying potential describes the energetics of the ground state—the planar state in the

valley, and the twisted state on the wall sides. After ionization, the upper deep potential valley is

reached, either in the stable state (black point) or in the twisted situation placed at the walls. The

metastable radical cation (block circle) can undergo delayed fragmentation, which is an exergonic

process. The twisted radical cation on the top of the wall (open circle), however, will undergo an

immediate

endergonic fragmentation. Some arguments support this hypothesis:

1. Kinetic control instead of the thermodynamic one: the vibration motion of the critical bond

is more than one order of magnitude faster than the rotation (Table 10.3). Hence, dissocia-

tion

happens within the rst vibration motion and yields unfavorable products.

2. The cation afnity of the solvent BuCl decreases the dissociation energy compared to

gas-phase

conditions, which are used for calculations.

3. FET is a very exergonic process, which also causes vibration excitation.

↺

400 500 600 700

0.00

0.05

0.10

1 2

0.00

0.02

0.04

t (μs)

335 nm

1 2

0.00

0.02

0.04

420 nm

–

•

CPh

ΔOD

λ (nm)

2-NpS

•

(a)

(b)

Figure 10.10 Transient optical absorption spectra obtained in the pulse radiolysis of a N

-saturated

solution of 2-NpSCPh

(2 × 10

−3

mol dm

−3

) in BuCl. Spectra correspond to times (◽) 45ns, (▵) 140 ns, and

(◦)700ns after the pulse. Experimental time proles demonstrating the inuence of (a) Bu

−

and (b) O

the kinetic behavior of the transients are given in insets.

Reactivity of Radical Cations in Nonpolar Condensed Matter 253

Considering all of these factors, an immediate fragmentation can explain the energetically unfavor-

able products, such as NpS

and

•

CPh

. Clearly, rapid relaxation (reorganization) to the metastable

radical cation can also happen. Because of the excess energy, this would result in an immediate

dissociation in the energetically favorable channel, forming NpS

•

and

CPh

, in (10.14b) given in

brackets. Hence, the yield of the unusual dissociation products does not give a quantitative measure,

but

is a crucial information about the FET mechanism.

10.4.5 driving force of fet

The energetic aspects of the dissociation of the unstable donor radical cations (reaction path 10.9b)

have already been discussed. Now, the dependence of the ET on the free energy of the gross reaction

(10.4c) will be analyzed. Here, we use the difference between the IPs of the nonpolar solvent and the

donor (ΔIP) as a measure of ΔG. In earlier ET studies, we observed the expected dependence of the

rate

constants on ΔIP (see Figure 10.12) (Mehnert etal., 1985).

The

investigations were performed without product control, because only the inuence of the added

donors on the time proles of the primary solvent radical cations was evaluated to determine rate con-

stants. The rate constants were normalized on the diffusion-controlled limit for the particular solvent.

Although interpreted with another theoretical model (Brede etal., 1987), the plots show the typical

shape of the well-known Marcus plots (Eberson, 1982). This macroscopic behavior can be understood

in terms of the usual sum kinetics. From the plot, it can be seen that the ET shows activation-controlled

rate constants up to ΔIP ≤ 0.3eV. Then the diffusion-controlled plateau is reached.

Now, we vary the ΔIP values in another way, that is, by stepwise ET. The concept is that primary

stable aromatic radical cations should be formed, which could then be used as electron acceptors for

aromatic sulde. This is shown in the reaction sequence (10.15):

0 90

Electron jump

Rotation degree

ground state

radical cation

Energy (kJ mol

–1

)

Planar

Twisted

Fragmentation

Reorganization

Fragmentation

2-NpS

+ CPh

2-NpS +

CPh

180

•

Figure 10.11 Potential energy diagram describing energetic proles of the planar and twisted radical

cation

conformers and subsequent fragmentation reactions.

254 Charged Particle and Photon Interactions with Matter

BuCl C H BuCl C H

6 6 6 6

i i+ +

+ → + (10.15a)

C H Ar-S-CPh C H

6 6 3 6 6

+ → + products

(10.15b)

So, the ΔIP value has been varied from about 3eV down to 0.45eV. As a criterion for judging the reac-

tion course, we looked for the product pattern derived from the sulde, that is, whether trityl cations as

well as trityl radicals were formed (Karakostas etal., 2007). This concept is illustrated in Figure 10.13.

As a result of the ET from the sulde to the radical cations of butyl chloride, benzene, or butyl-

benzene, both trityl species were found in each case. The ET from biphenyl to the sulde, however,

resulted only in trityl cations. This led to a conclusion that in the rst case, the FET took place with

an electron jump occurring immediately after the rst collision of the partners. In the second case,

it is hypothesized that in the ion–molecule reaction, an encounter complex has been passed, and,

as a consequence, only the way via the metastable intermediate sulde radical cation was followed

(Karakostas etal., 2007; Baidak etal., 2008b). Here, metastable refers to a short-lived species, but

not a dissociative one, which would be formed in femtoseconds. These ndings are also in accord

with stepwise ET experiments using phenol as the nal donor and monitor (Brede etal., 2003).

One can conclude that at ΔIP below 0.5eV, aprocess similar to a classical diffusion-controlled ET

happens,

that is, an ET with a high rate but within an encounter complex situation.

summary, the following can be drawn: (a) FET with its typical product situation takes place at

free energies of the gross reaction of −ΔG higher than 0.5eV. (b) Below this value, the gross reaction

CHCl

i-C

n-C

c-C

n-C

3-Me-heptane

2-Me-hexene-1

3-Me-

cyclohexene

Toluene

Heptene-1

Hexene-2

III

ΔJ (eV)

k (dm

mol

–1

)

2 3

Hexene-1

Figure 10.12 Dependence of the rate constant, k

, on ΔIP. The donors are indicated. Electron acceptors

were different alkane or alkyl chloride parent radical cations. Solid and dotted lines are calculated for differ-

ent

deformation energies, as described by Brede etal. (1987).

Reactivity of Radical Cations in Nonpolar Condensed Matter 255

also exhibits diffusion-controlled rate constants (k ≈ 10

mol

−1

), but seems to proceed via

an encounter complex. (c) At energies of −ΔG below 0.3eV, activation-controlled rate constants are

observed.

10.4.6 relation between Molecule dynaMicS and reaction kineticS

The dynamic electron shifts in the donor orbitals documented above (Figures 10.7 and 10.9) do not

play a role in common bimolecular chemical reactions. In normal ET reactions, the timescale is

usually longer than a nanosecond (this is the time range of diffusion). However, in the exotic case,

where the reaction-determining step is faster than the bending (rotation) motions, the discrete con-

formations of the deformed molecules in their momentary states strongly inuence the reaction.

This can be experimentally observed, because different products are formed for different encoun-

ters. It involves the assumption that the deciding step of the reaction happens in the rst collision of

the

reaction partners, which also means that no dened encounter complex is passed.

Generally,

an encounter complex is dened as an entity formed after the approach of the reaction

partners, in which a more or less rigid and energetically optimized structure is reached (Weller,

1982; Klessinger and Michl, 1995; Rosokha and Kochi, 2008). In this structure, for common ET

processes (one-electron oxidation, photoinduced triplet- or singlet-sensitized ET, proton-coupled

ET, etc.), a process similar to a time-delayed, but rapid, electron jump from the donor part to the

acceptor happens (Klessinger and Michl, 1995). These ET processes follow the usual sum and equi-

librium

kinetics placed in the time range ≥ 1

ns.

For the analysis of FET, however, a dynamically controlled regime has to be used instead of the

mentioned description as an elementary reaction with thermodynamic parameters. Understanding

Free electron

transfer

Reaction-

controlled electron

transfer

10.84 eV

9.25 eV

8.69 eV

8.34 eV

Ionization potential (eV)

~7.9 eV

Intermediary

electron donors

Figure 10.13 The scheme symbolizes the ET under the variation of free energy of the reaction by stepwise

(see reactions 10.15a and 10.15b).

256 Charged Particle and Photon Interactions with Matter

the FET as a reaction of a diversity of (rotational) conformers (instead of one uniform molecule

type), which by the electron jump forms a diversity of ionization products, replaces the commonly

used picture of mesomerism (Beyer and Walter, 1991) of substituted aromatic molecules in this case.

If at all such a mesomerism exists, then it is a property of the encounter complex, for which a rigid

and

energetically favored structure should exist.

The

main feature of FET consists in the fact that the femtosecond dynamic motions of (donor)

molecules are reected in a diffusion-controlled reaction taking place at the nanosecond tim-

escale. Therefore, with the used time-resolved spectroscopy, a steady state observation of the

transients, which appear in nanoseconds (see bimolecular reaction), is performed. However,

their structure is determined by femtosecond dynamics. Insofar, the FET is a new phenomenon

of reaction kinetics.

10.4.7 coMpariSon of et theorieS

Here, the classical ET models are sketched briey, mainly under the aspect of the distribution of

excess

energy of ET.

already mentioned, common interpretations of ET reactions are based on sum kinetics and use

equilibria for the energetic description (cf. reaction 10.8). To describe the dependence of the ET trans-

fer rate constant, k

(rate of the gross reaction), on the free energy change, ΔG

, two empirical attempts

formulated by Marcus (Marcus, 1964; Marcus and Sutin, 1964) and alternatively by Rehm and Weller

(1969) are used. In both models, the excess energy is distributed either within the reactant molecules or

in the solvent surroundings. In polar media, this reorganization energy concerns primarily the solvent

rearrangement, taking into account mainly excitation of the low-energy modes of the solvent.

In contrast to the situation discussed above, for nonpolar media, the solvent reorganization energy

is expected to play a minor role, and, hence, the reorganization of high-frequency intramolecular

vibrational modes will determine the ET rate. Taking into account the quantum nature of these

vibrations, the description of k

was modied. Using the concept of Ulstrup and Jortner (1975), the

ET in alkanes and alkyl chlorides (Brede etal., 1987) has been described. Taking C-H modes of the

solvent radical cation/solute couple, and treating the solvent part in a classical way as proposed by

Levich

and Dogonadze (1961), an expression for k

could be obtained.

In principle, the FET phenomenon can use parts of both theories discussed above. However,

the theory developed for nonpolar systems is more appropriate (Brede etal., 1987). It should be

noted that in FET, molecular dynamics plays a decisive role (Brede and Naumov, 2006) com-

pared to the classical descriptions of ET, which are based on a kinetic analysis using thermo-

dynamic data. The FET process, however, can only be understood in terms of intramolecular

mobility (vibrations and bending motions) of the donor. With the peculiarity of a prompt electron

jump taking place in the rst approach of the reaction partners, instead of one type of donor mol-

ecules, we have to take into account a dynamic diversity of conformers with different electron

distributions in the momentum of the electron jump. This leads to the distinction of two types of

rotation conformers. These conformers react with solvent radical cations, forming at rst donor

radical cations of different stability (dissociative and metastable), which nally result in different

products (fragments and metastable radical cations). Most of these events occur in the femtosec-

ond time domain (dissociation), or even faster (electron jump). Considering bimolecular FET, a

kinetic paradox is observed, that is, instead of the slowest step, the fastest one determines the

reaction path.

Overall, FET is an electron transfer process that is governed by molecular dynamics. Apart from

the unusual “two-product” situation, it is also noteworthy that the endergonic fragmentation prod-

ucts derived from the dissociative radical cations do not follow the rules of the classical energetic

analysis

(Baidak etal., 2008a).

illustrate the discussion on the time range of FET and common ETs, in Figure 10.14, relevant

processes are assigned to a logarithmic timescale. This should visualize the crucial distinction

Reactivity of Radical Cations in Nonpolar Condensed Matter 257

between ET sum kinetics under equilibrium conditions on one side and the dynamically controlled

FET on the other side. Because both ET processes are bimolecular, the observation is placed in the

nanosecond real-time spectroscopy, which is an adequate technique for studying such diffusion-

controlled reactions. Concerning the dynamics of the femtosecond events mentioned above, the

investigations

were carried out under quasi steady state aspects.

10.5 transFormation oF donor radiCal Cations in Freon matriCes

10.5.1 reaction principleS

Rapid transformations of donor radical cations formed by ET to freon parent ions (reaction 10.3c)

can be observed by matrix insulation studies in electron-irradiated frozen freon matrices at low

temperature. Under these conditions, FET cannot take place because of the very limited mobility

of the donor molecule. For this purpose, low-temperature EPR measurements in combination with

quantum chemical calculations were performed (Knolle etal., 1999; Janovský et al., 2003; Naumov

etal., 2003a,b,c,d). Depending on the experimental conditions, the transformation of the primary

radical cations proceeds via monomolecular or bimolecular reactions. At high solute concentrations

and in the CFCl

Cl (freon F-13) matrix, the transformations often happen by intermolecular

H-abstraction or proton transfer from the parent molecule, forming a deprotonated radical and a car-

bocation, or performing an addition reaction, which leads to a dimer radical cation (Naumov etal.,

2003b; Janovský etal., 2005). The dimer radical cation, which could be a distonic one, character-

ized by separation of charge and radical site, can react further through intermolecular H-abstraction

from the parent molecule, forming a dimer carbocation and a radical (Naumov etal., 2003c, 2005).

Subsequently, the dimer radical cation can perform an addition reaction with the parent molecule,

forming

a trimer radical cation, etc.

10.5.2 intraMolecular tranSforMationS

At low solute concentrations and using a CF

CCl

matrix (which generally does not form complexes

with the solute cations), the transformation can proceed in a monomolecular way, such as intramo-

lecular H-shift, ring closing (opening), conformational changes, and decomposition. Here, some

examples of monomolecular rearrangements of donor radical cations by intramolecular H-shift and

ring

closing are demonstrated.

The

reaction pathways of experimentally observed intramolecular transformation of the primary

radical cations of 2,5-dihydrofuran (2,5-DHF) (Janovský etal., 1999; Naumov etal., 2003a,d) and

Dynamic control

Classical ET

(equilibrium kinetics)

FET gross reaction

Diﬀusione-jump

Vibration

(stretching)

Bending vibration

(rotation motion)

–15 –12 –9 –6 Log (t)

fs ps ns

Figure 10.14 Time correlation of the ET processes discussed in this chapter.

258 Charged Particle and Photon Interactions with Matter

of 2,5-dihydropyrrole (2,5-DHP) (Naumov etal., 2004a) are shown in Figure 10.15. The primary

oxygen-centered 2,5-DHF

+•

transforms into 2,4-DHF

+•

, which is stable at 77 K. A subsequent transfor-

mation of 2,4-DHF

+•

can be induced by illumination with visible light. This process is an intramolecu-

lar H-shift, leading to the formation of 2,3-DHF

+•

and 3,4-DHF

+•

, the latter being a newly identied

species, stable in the range of 77–145K, characterized by hfs (hyperne splitting constants) a(2H) =

1.59mT and a(4H) = 2.82mT.

This transformation pattern was observed also in the case of 2,5-DHP

+•

(Figure 10.15b). The

primary nitrogen-centered radical cation, 2,5-DHP

+•

, in freon matrices is relatively stable (in

CCl

up to 140 K); however, on illumination with visible light of wavelength λ > 600 nm, it is

transformed into 2,4-DHP

+•

by an intramolecular 5 → 4 H-shift. The latter is stable up to 143K,

but its further intramolecular rearrangements through 2 → 3 and 4 → 3 H-shifts can be induced by

illumination with visible light of shorter wavelengths (400nm < λ < 600nm). Both transformations

proceed simultaneously, with the same yields of the two isomers of the DHP radical cation, namely,

2,3-DHP

+•

and 3,4-DHP

+•

. These transformations are most feasible also due to the strong localiza-

tion

of the unpaired electron in position 4 in 2,4-DHP

+•

The thermal and photochemical transformations of primary amine radical cations generated

radiolytically in freon matrices have been comprehensively investigated using low-temperature EPR

spectroscopy (Janovský etal., 2004; Knolle etal., 2006). The primary nitrogen-centered amine

radical cations are very unstable even at 77K and undergo a monomolecular transformation through

an H-shift into a more stable structure (Figure 10.16). It should be noted that, in agreement with the

calculated high activation energy, a rearrangement of the primary nitrogen-centered amine radical

cations by a hypothetical H-atom shift from C1 to the ionized NH

group was not observed. The

H-shift

goes in the direction of the lower barrier.

HC CH

CH2

2-3 H-shift

2,4-DHF

•

2,5-DHF

•

3,4-DHF

•

2,3-DHF

•

∆H(2–3) = –6

(2–3) = 67

5-4 H-shift

3-4 H-shift

∆H(2–3) = –17

(2–3) = 124

∆H(2–3) = –68

(2–3) = 75

2-3 H-shift

2,4-DHP

•

2,5-DHP

•

3,4-DHP

•

2,3-DHP

•

∆H(2 – 3) = –23

(2 – 3) = 90

5-4 H-shift

3-4 H-shift

∆H(2–3) = –20

(2–3) = 132

∆H(2–3) = –44

(2–3) = 109

C H

(a)

(b)

•

Figure 10.15 Experimentally observed intramolecular transformations of primary (a) 2,5-DHF

•

and

(b)2,5-DHP

•

through the H-shift within the molecular ring. ΔH and E

are the calculated reaction enthalpies

and

the activation energies (kJ mol

−1

), respectively, for the H-shift.

Reactivity of Radical Cations in Nonpolar Condensed Matter 259

A further monomolecular rearrangement of a primary donor radical cation was studied in the

case of lactones of different sizes (Naumov etal., 2004b). While the primary radical cation of a four-

membered ring β-butyrolactone is unstable at 77K and undergoes ring opening and fragmentation,

ending at the deprotonated neutral (CH

CHCH

)

•

radical, the primary radical cations of the ve-,

six-, and seven-membered ring lactones could be directly observed. Here, the transformation of the

primary carbonyl-centered radical cation proceeds by an intramolecular H-shift to the carbonyl

oxygen (Figure 10.17). Calculations as well as experiments show that the H-shift occurs always

from C1 in the α-position to the carbonyl group. The stability of the primary radical cation toward

transformation depends on the ring size, which determines the activation energy of the transforma-

tion and the distance, L(H–O), of the carbonyl oxygen from the nearest H-atom on the ring. The

larger the ring, the smaller the L(H–O) and also the activation energy of the H-shift, making the

transformation

of the primary radical cation more feasible.

The

transformations of radical cations of ethyl acrylate (EtA) radiolytically generated in freon

(CFC

Cl, CF

, and CFC

) and in argon matrices were investigated using low-temperature

EPR spectroscopy (Knolle etal., 2002). The primary cation of EtA could be trapped below 40 K

in a CFCl

matrix only. In all the freon matrices studied, the intramolecular rearrangement of this

cation occurs via the H-shift from the CH

group to carbonyl oxygen, leading to the formation of a

distonic

radical cation (Figure 10.18).

Although

the calculated reaction enthalpy for the H-shift from the CH

group to carbonyl oxygen

is larger than that for the H-shift from the CH

group, the reaction proceeds in the direction of the

lower energetic barrier, apparently due to the more favorable six-ring transition structure. In the case

of the CF

CCl

matrix, warming the samples to temperatures above 130K results in the simulta-

neous formation of two new species, which were assigned to six- and ve-member ring structures

(a(H)/mT: 2.36 (Hα), 5.13 (Hβ1), 1.9 (Hβ2), and 2.27 (2Hα) and 2.6 (Hβ), respectively) formed by

intramolecular

cycloaddition (ring closing) of the terminal radical to the vinyl double bond.

Propylamine

•

+•

ΔH = –11

ΔH = –15

= 72

= 177

= 169

ΔH = 4

Allylamine

ΔH = –49

ΔH = –86

= 43

= 70

= 109

ΔH = –121

Propargilamine

ΔH = –105

ΔH = –46

= 61

= 143

= 119

ΔH = –61

•

Figure 10.16 Experimentally observed intramolecular transformations of amine radical cations through the

intramolecular H-shift. ΔH and E

are the calculated reaction enthalpy and the activation energy (kJ mol

−1

respectively, for the H-shift.