**4. Specific features of nitroxide and peroxide radicals reactions in biological systems and in liquid phase organic substrates**

Like transition metals, nitroxide radicals can easily transform both to oxidized (oxoammonium cation) and to reduced (hydroxylamine) forms (Berliner, 1998; Sen' & Golubev, 2009; Zhdanov, 1992). This fact, along with >NO• ability to penetrate through cell membranes and with its paramagnetic properties, suggests that nitroxide radicals poses a number of unique features unlike the compounds of any other class. Like antioxidants and mimetics of superoxide dismutase enzyme, >NO• are able to supply an effective protection for cells and tissues (Denisov & Afanas'ev, 2005; Soule et al., 2007; Wilcox, 2010). Nitroxide radicals inhibit oxidation of lipids in fatty acids' micelles (Damiani et al., 2002; Noguchi et al., 1999), liposome membranes (Damiani et al., 2002; Wilcox, 2010), lipoproteins (Damiani et al., 1994), and in low density microsomes (Antosiewicz et al., 1995; Nilsson et al., 1990). At the moment there's not enough information concerning >NO• reactions with redox peroxide

<sup>2</sup> Inhibiting factor is ratio of real induction period (*t*) to theoretical period of inhibitor conversion (), i.e. *f*inh = *t*/, where = *f*[>NO•]/*W*i (*f* – stoichiometric inhibiting factor).

radicals to understand such reactions' mechanisms. In fact, there is no any certain agreement on this reaction, so some authors consider it as unlikely one (Blough, 1988; Browlie & Ingold, 1967; Damiani et al., 2002; Denisov, 1996).

Our recent results concerning reactions of >NO•(I) – (VII) with MO2• radicals of 1,1 substituted ethylenes allows suggestion that such reaction's probability is quite high.

Let's review some experimental data in this field to understand the situation.

270 Nitroxides – Theory, Experiment and Applications

*i*

where *k*5 = (*k*5.1[>NO•] + *k*5.2[>NOH])/2[>NO•]0.

Aleksandrov, 1977; Ruban et al., 1967).

than ten and reflects just a lower bound of this value.

where = *f*[>NO•]/*W*i (*f* – stoichiometric inhibiting factor).

1996):

reaction (4) to chain termination process is negligible, and then

• • 0.5

0.5

*i*

(4)

(3)

0 40 05 0

[O ]

Kinetic analysis shows that at [O2] ~ 1∙10–2 M and [>NO•]0 < 10–4 M the contribution of

0 50

*W k W W W W k*

<sup>0</sup> <sup>3</sup>

With the drop of *P*o2 and small share of quadratic chain termination the oxidation rate will decrease not linearly, but slower. Such facts were found for instance in (Pliss &

Reaction (5.1) proceeds as disproportionation of nitroxide and peroxide radicals (Denisov,

HO2• + >NO• >NOH + O2

>C(OH)O2• + >NO• >NOH + >C=O +O2

>CH–CH(OO•)N< + >NO• >NOH + >C=CH–N< + O2 >NO• regeneration and multiple chain termination processes are caused just by subsequent reaction (5.2). Measured kinetic inhibiting factors2 for different nitroxide radicals and substrates presented for example in review (Denisov, 1996). The most of *f* values greater

Like transition metals, nitroxide radicals can easily transform both to oxidized (oxoammonium cation) and to reduced (hydroxylamine) forms (Berliner, 1998; Sen' & Golubev, 2009; Zhdanov, 1992). This fact, along with >NO• ability to penetrate through cell membranes and with its paramagnetic properties, suggests that nitroxide radicals poses a number of unique features unlike the compounds of any other class. Like antioxidants and mimetics of superoxide dismutase enzyme, >NO• are able to supply an effective protection for cells and tissues (Denisov & Afanas'ev, 2005; Soule et al., 2007; Wilcox, 2010). Nitroxide radicals inhibit oxidation of lipids in fatty acids' micelles (Damiani et al., 2002; Noguchi et al., 1999), liposome membranes (Damiani et al., 2002; Wilcox, 2010), lipoproteins (Damiani et al., 1994), and in low density microsomes (Antosiewicz et al., 1995; Nilsson et al., 1990). At the moment there's not enough information concerning >NO• reactions with redox peroxide

<sup>2</sup> Inhibiting factor is ratio of real induction period (*t*) to theoretical period of inhibitor conversion (), i.e. *f*inh = *t*/,

**4. Specific features of nitroxide and peroxide radicals reactions in** 

**biological systems and in liquid phase organic substrates** 

*i*

2 [>NO ]

•

0.5

*W k Wk W <sup>W</sup> <sup>W</sup> WW k k*

0 12 3

[>NO ] 2 [>NO ]

### **4.1. Nitroxide and peroxide radicals reaction under conditions modeling biological systems**

Barton et al. assumed that in reaction of >NO• (I) with (CH3)3COO• the formation of quite stable intermediate should take place (Barton et al., 1998):

$$\text{\color{red}{\cdot \text{NO}^\*}}\text{\color{red}{\cdot \text{(CH}\text{\cdot})}\text{\color{red}{\cdot \text{COO}^\*}}\text{\color{red}{\cdot \text{(CH}\text{\cdot})}\text{\color{red}{\cdot \text{(CH}\text{\cdot})}\text{\cdot}}\text{\color{red}{\cdot \text{(CH}\text{\cdot})}\text{\cdot}}$$

Further this intermediate should decompose to a number of products with concomitant formation of molecular oxygen and >NO• regeneration according to the following very speculative scheme (Scheme 4) with no any kinetic evidence (Barton et al., 1998).

**Scheme 4.** Mechanism of nitroxide with tert-butyl peroxide radicals reaction according to (Barton et al, 1998)

Stipa attempted to prove this scheme with *ab initio* quantum-chemical calculations (Stipa, 2001). The calculations were performed using Gaussian 98 Hartree-Fock method and complete basis set CBS-QB3. H2NO• Radical was used as a model of >NO• (I) in view of available computer resources limitations (work was submitted in April 2001). The calculations' results indicate the possibility of Scheme 4.

This conclusion seems to be quite controversial since radical H2NO• can not provide an adequate >NO• (I) model. For example, standard enthalpy calculated for the first reaction

stage was –1.794 kcal/mol (tabelle 2 in (Stipa, 2001)), i.e. this value is not very different from zero within the quantum-chemical calculations accuracy. This suggests that the studied trioxide apparently must be thermodynamically unstable structure. At the same time, radical H2NO• is much more active than simulated >NO• (I). It's easy to show with Density Functional Theory (DFT) calculations (Becke, 1993). Quantum energy values of radicals H2NO• and >NO• (I) recombination with CH3O• and CH3OO• were calculated by DFT B3LYP/6-31G\* (similar calculation is shown in Table 3). The results are shown below.


#### **Table 4.**

Even this simple example shows that stable nitroxide radical >NO• (I) is significantly less active than H2NO•.

Offer and Samuni also suggest that stable trioxide formation proceeds in reaction of tertamidinopropyl radicals with >NO• (I) and (III) in phosphate buffer at pH 7.4 and 310 K (Offer & Samuni, 2002):

$$\text{(HzN)} \natural \text{CC(CHz)CO}^{\bullet} + \text{>NO}^{\bullet} \longrightarrow \text{(HzN)} \natural \text{CC(CHz)COON} < \text{}$$

This reaction was studied with combination of ESR-spectroscopy method and cyclic voltammetry, but no any kinetic evidences of product formation and its subsequent transformation were provided.

Brede et al. studied the reaction of *n*-C17N35OO• with >NO• (I) with pulse radiolysis at room temperature (Brede et al., 1998). The rate constant value *k* < 105 M–1∙s–1 was specified but the reaction's mechanism wasn't discussed.

Goldstein S. and Samuni also determined the rate constants of reactions of CH3OO•, (CH3)3COO•, CH2(OH)OO•, and HO2• with >NO• (I) – (IV), (X) – (XII) with pulse radiolysis at room temperature (Goldstein & Samuni, 2007). The values obtained are shown in Table 4. The mechanism of RO2• + >NO• reaction was discussed; therefore let's consider this paper in more detail.

It was shown that nitroxide with peroxide radicals' reaction mainly results in formation of corresponding oxoammonium cations. In case of (CH3)3COO• the formation of stable cation radical >NO• (I) and decay products of relatively unstable cation radical >NO• (II) was spectrophotometrically detected. In case of HOO• the reaction with piperidine nitroxide radicals is catalyzed by anion H2PO4– as noted in (Goldstein & Samuni, 2007). Therefore under physiological conditions (pH 7.4 and 5∙10–2 M phosphate) observed rate constants for piperidine nitroxide radicals are slightly greater than those shown in Table 4. Moreover, catalysis by H2PO4– anions implies that nitroxide with peroxide radicals' reaction proceeds according to inner-sphere electron transfer mechanism when >NOOOR adduct's decomposition can undergo general acid catalysis:

272 Nitroxides – Theory, Experiment and Applications

**Table 4.**

active than H2NO•.

more detail.

(Offer & Samuni, 2002):

transformation were provided.

reaction's mechanism wasn't discussed.

stage was –1.794 kcal/mol (tabelle 2 in (Stipa, 2001)), i.e. this value is not very different from zero within the quantum-chemical calculations accuracy. This suggests that the studied trioxide apparently must be thermodynamically unstable structure. At the same time, radical H2NO• is much more active than simulated >NO• (I). It's easy to show with Density Functional Theory (DFT) calculations (Becke, 1993). Quantum energy values of radicals H2NO• and >NO• (I) recombination with CH3O• and CH3OO• were calculated by DFT

Radical Recombination energy, kJ/mol

Even this simple example shows that stable nitroxide radical >NO• (I) is significantly less

Offer and Samuni also suggest that stable trioxide formation proceeds in reaction of tertamidinopropyl radicals with >NO• (I) and (III) in phosphate buffer at pH 7.4 and 310 K

(H2N)2+CC(CH3)OO• + >NO• (H2N)2+CC(CH3)OOON This reaction was studied with combination of ESR-spectroscopy method and cyclic voltammetry, but no any kinetic evidences of product formation and its subsequent

Brede et al. studied the reaction of *n*-C17N35OO• with >NO• (I) with pulse radiolysis at room temperature (Brede et al., 1998). The rate constant value *k* < 105 M–1∙s–1 was specified but the

Goldstein S. and Samuni also determined the rate constants of reactions of CH3OO•, (CH3)3COO•, CH2(OH)OO•, and HO2• with >NO• (I) – (IV), (X) – (XII) with pulse radiolysis at room temperature (Goldstein & Samuni, 2007). The values obtained are shown in Table 4. The mechanism of RO2• + >NO• reaction was discussed; therefore let's consider this paper in

It was shown that nitroxide with peroxide radicals' reaction mainly results in formation of corresponding oxoammonium cations. In case of (CH3)3COO• the formation of stable cation radical >NO• (I) and decay products of relatively unstable cation radical >NO• (II) was spectrophotometrically detected. In case of HOO• the reaction with piperidine nitroxide radicals is catalyzed by anion H2PO4– as noted in (Goldstein & Samuni, 2007). Therefore under physiological conditions (pH 7.4 and 5∙10–2 M phosphate) observed rate constants for piperidine nitroxide radicals are slightly greater than those shown in Table 4. Moreover,

HO• –28.3 –98.4 HOO• 48.1 –20.4 CH3• –150.6 –232.8 CH3O• 11.5 –58.2 CH3OO• 60.3 9.92

>NO•(I) H2NO•

B3LYP/6-31G\* (similar calculation is shown in Table 3). The results are shown below.

$$\begin{aligned} \text{>NO}^{\bullet} + \text{ROO}^{\bullet} &\xleftarrow{\longrightarrow} \text{>NOOOR} \\\\ > \text{NOOOR} &\xleftarrow{\longrightarrow} \text{N}^{+} = \text{O} + \text{ROO}^{\bullet} \\\\ > \text{NOOOR} &+ \text{H}\_{2}\text{PO}\_{4}^{-} &\xleftarrow{\longrightarrow} \text{N}^{+} = \text{O} + \text{ROOH} + \text{HPO}\_{4}^{2-} \end{aligned}$$

It's obvious that in absence of catalysis the adduct's formation may also occur, but innersphere electron transfer mechanism still can not be excluded:

$$\text{>NO}^\* + \text{ROO}^\* \rightleftharpoons \text{N}^\* = \text{O} + \text{ROO}^-$$

Known rate constants of reaction HO2• (>C(OH)O2•) + >NO• (I) and (III) in organic solvents are in the range of 1.1∙104 – 2.1∙105 M–1∙s–1 at 323 K (Aleksandrov, 1987; Wilcox, 2010), which is two-three orders of magnitude lower than values shown in Table 5. Such a huge difference may hardly be explained only by the reaction's specificity in phosphate buffer. Further reaction of radicals HO2• and CH2(OH)OO• with >NO• proceeds, as is well known, as disproportionation rather than recombination (Denisov, 1996; Denisov & Afanas'ev, 2005; Mogilevich & Pliss, 1990).


\*) Calculated on the base of (Goldstein et al., 2003; Goldstein & Samuni, 2007).

**Table 5.** Rate constants (M–1∙s–1) of ROO• + >NO• reaction (Goldstein & Samuni, 2007).

It's important that authors (Goldstein & Samuni, 2007) on the base of experimental data suggested that rate constant of reaction RO2• + >NOH doesn't exceed 1∙105 M–1∙s–1, i.e. this reaction is much slower as compared with RO2• + >NO• (Table 4). Thereby we note that, as it well known (Denisov, 1996; Denisov & Afanas'ev, 2005; Mogilevich & Pliss, 1990), upon a competition of reactions (5.1) and (5.2) (see Scheme 2) the rate-limiting reaction is just the first one.

As can be seen from the results above, the mechanism of nitroxide with peroxide radicals' reactions in biological systems isn't elucidated at all. The authors of works analyzed are inclined to possibility of >NO• reaction not only with peroxide radicals poses redox

properties: HO2•, >C(OH)O2•, but with RO2• which are oxidants only: (CH3)3COO•, *n*-C17H35OO• (Barton et al., 1998; Brede et al., 1998; Goldstein et al., 2003; Goldstein & Samuni, 2007; Offer & Samuni, 2002 ; Stipa, 2001). It is assumed that such a reaction occurs as recombination with formation of trioxide. According to the authors of papers cited, nitroxide radicals are able to provide protection against oxidation at extremely low concentrations due to the regeneration process resulting from the reaction of corresponding oxoammonium cations with common biological reducing agents.

Hereby it's interesting to consider the reaction of >NO• with RO2• which are only oxidants under conditions when oxoammonium cations' formation is improbable, i.e. during an oxidation in organic phase.

### **4.2. Nitroxide radicals reactions with peroxide radicals of 1- and 1,1- ethylene substituted monomers**

As it has already been mentioned above, reactions of >NO• with peroxide radicals of 1- and 1,1- ethylene substituted monomers were discovered recently (Pliss et al., 2010a, 2010b, 2012). Authors (Pliss et al., 2010a, 2010b) on the basis of kinetic data of styrene's and (meth)acrylates' oxidation in presence of piperidine, pyrroline, and imidazoline >NO• suggested that nitroxide radicals inhibit the oxidation process via >NO• reaction with substrate's both alkyl and peroxide radicals, and moreover >NO• regeneration occurs during chain termination process in accordance with reactions:

$$\text{MOr}^\* + \text{>NO}^\* \longrightarrow \text{product} + \text{>NOH} \tag{5.1}$$

$$\text{MOr}^{\bullet} + \text{>NOH} \longrightarrow \text{MOOH} + \text{>NO}^{\bullet} \tag{5.2}$$

The following results served grounds for these assumptions: oxygen consumes linearly in presence of >NO• for all monomers for a long period of time. That period is greater than the theoretical induction period, and if *P*o2 reduces to five times (from 1∙105 to 0.2∙105 Pa) the oxidation rate decreases to less than two times (Pliss et al., 2010a, 2010b). But of course it is unacceptable to make any definite conclusions about detailed mechanism just on the basis of the oxygen consumption kinetic data only.

In (Pliss et al., 2012) >NO• antioxidant activity was studied during styrene's oxidation using a complex of kinetic methods in combination with quantum-chemical calculations and kinetic modeling. The choice of styrene is caused by the following circumstances. Firstly, the reaction of styrene's inhibited oxidation is not complicated by complexation process as it is in case of many other vinyl compounds (acrylic monomers for example). Secondly, high reactivity of styrene's double bond makes it possible to study this process under long chains conditions even if oxidation is quite strongly inhibited. Thirdly, in case of styrene the rate constants of elementary stages are known for many key reactions, and this knowledge makes kinetic modeling significantly easer to carry out.

Let's consider the results of this work and some of our new data.

Oxidation kinetics was studied in area of initial O2 consumption rates in temperature range of 310 – 343 K with highly sensitive capillary microvolumometer according to technique (Loshadkin et al., 2002). Initial >NO• concentrations were in range of 10–7 – 10–3 M. Experiments were carried out at *P*o2 = 20 or 100 kPa. In special cases oxygen-argon mixes were prepared to obtain the oxidation rates dependences on [O2]. Initiation rate *W*i was determined with inhibition method by detection of induction period ending time ind and application of known equation *W*<sup>i</sup> = 2[InH]0/ind. 6-Hydroxy-2,2,5,7,8-pentamethylbenzochroman was used as inhibitor (InH). Kinetic modeling was performed as described in (Loshadkin et al., 2002).

274 Nitroxides – Theory, Experiment and Applications

oxidation in organic phase.

**substituted monomers** 

properties: HO2•, >C(OH)O2•, but with RO2• which are oxidants only: (CH3)3COO•, *n*-C17H35OO• (Barton et al., 1998; Brede et al., 1998; Goldstein et al., 2003; Goldstein & Samuni, 2007; Offer & Samuni, 2002 ; Stipa, 2001). It is assumed that such a reaction occurs as recombination with formation of trioxide. According to the authors of papers cited, nitroxide radicals are able to provide protection against oxidation at extremely low concentrations due to the regeneration process resulting from the reaction of corresponding

Hereby it's interesting to consider the reaction of >NO• with RO2• which are only oxidants under conditions when oxoammonium cations' formation is improbable, i.e. during an

**4.2. Nitroxide radicals reactions with peroxide radicals of 1- and 1,1- ethylene** 

As it has already been mentioned above, reactions of >NO• with peroxide radicals of 1- and 1,1- ethylene substituted monomers were discovered recently (Pliss et al., 2010a, 2010b, 2012). Authors (Pliss et al., 2010a, 2010b) on the basis of kinetic data of styrene's and (meth)acrylates' oxidation in presence of piperidine, pyrroline, and imidazoline >NO• suggested that nitroxide radicals inhibit the oxidation process via >NO• reaction with substrate's both alkyl and peroxide radicals, and moreover >NO• regeneration occurs

MO2• + >NO• product + >NOH (5.1)

MO2• + >NOH MOOH + >NO• (5.2)

The following results served grounds for these assumptions: oxygen consumes linearly in presence of >NO• for all monomers for a long period of time. That period is greater than the theoretical induction period, and if *P*o2 reduces to five times (from 1∙105 to 0.2∙105 Pa) the oxidation rate decreases to less than two times (Pliss et al., 2010a, 2010b). But of course it is unacceptable to make any definite conclusions about detailed mechanism just on the basis of

In (Pliss et al., 2012) >NO• antioxidant activity was studied during styrene's oxidation using a complex of kinetic methods in combination with quantum-chemical calculations and kinetic modeling. The choice of styrene is caused by the following circumstances. Firstly, the reaction of styrene's inhibited oxidation is not complicated by complexation process as it is in case of many other vinyl compounds (acrylic monomers for example). Secondly, high reactivity of styrene's double bond makes it possible to study this process under long chains conditions even if oxidation is quite strongly inhibited. Thirdly, in case of styrene the rate constants of elementary stages are known for many key reactions, and this knowledge

oxoammonium cations with common biological reducing agents.

during chain termination process in accordance with reactions:

the oxygen consumption kinetic data only.

makes kinetic modeling significantly easer to carry out.

Let's consider the results of this work and some of our new data.

Values of styrene's oxidation rates inhibited by different >NO• under oxygen and air saturation conditions and when quadratic termination share is no more than 25% are presented in Table 6. As seen from the table, the oxidation's rate for different >NO• from oxygen to air decreases substantially less than five times. This fact according to (Aleksandrov, 1987; Denisov & Afanas'ev, 2005; Kharitonov & Denisov, 1967; Kovtun et al., 1974; Mogilevich & Pliss, 1990; Pliss & Aleksandrov, 1977; Denisov, 1996) suggests that nitroxide radicals react with both M• and MO2•.

Kinetics of >NO•(III) consumption at its different initial concentrations are presented in Figure 2. From this data it follows that according to inhibition by reaction (4) >NO• should consume much faster than it happens in fact. With special experiments it was shown that value of inhibiting factor e.g. for >NO• (III) and >NO• (IV) is more than 7 (Pliss et al., 2012), so it suggests that >NO• regeneration process occurs upon chain termination. We note that this effect doesn't depend on styrene's concentration.


**Table 6.** Kinetic parameters of styrene's oxidation inhibited with >NO•; *W*i = 1.0∙10–8 M∙s–1 (Pliss et al., 2012)

**Figure 2.** Kinetics of >NO• (III) consumption during styrene oxidation (in relative coordinates); points: experimental data, curves: a result of simulation. *P*o2 = 20 kPa; *W*i = 1.0∙10–8 M∙s–1; [>NO•]0, M: 1 – 2.3∙10– 6, 2 – 5.9∙10–6, 3 – 1.2∙10–5, 4 – 2.8∙10–5 (Pliss et al., 2012).

Styrene's oxidation can be effectively inhibited by hydroxylamines. This assertion confirms with distinct induction period on oxygen consumption's kinetic curve (typical one represented at Figure 3 (Pliss et al., 2012)). Rate constants of such antioxidants' reactions with peroxide radicals can be determined from inhibited oxidation's rate dependence on time according to the following equation (Loshadkin et al., 2002):

$$F = \ln \frac{1 + \mathcal{W} / \mathcal{W}\_0}{1 - \mathcal{W} / \mathcal{W}\_0} - \frac{\mathcal{W}\_0}{\mathcal{W}} = \frac{k\_{5.2} \mathcal{W}\_0}{k\_2 \text{[M]}} t + const \tag{5}$$

But the problem is that oxidation rate remains significantly lower than *W*0 after the inhibiting period (see Figure 3) since >NO• being formed is an inhibitor of oxidation itself. However the kinetic modeling shows that in this case in equation (5) we can substitute *W*<sup>0</sup> value to the value of oxidation's rate at after-induction period. Herewith *k*5.2 determination error is less than 10%. Calculated *k*5.2 mean value at 323 K for >NOH (III), equal to 4∙106 M–1∙s–1 (Pliss et al., 2012), was later used in mechanism's kinetic modeling.

Dependences of >NO• consumption and its accumulation from corresponding >NOH on time during styrene's oxidation at *P*o2 = 20 kPa presented at Fig. 4 (ESR-spectroscopy method). It's seen that hydroxylamine being injected quickly transforms to >NO• and its maximum concentration differs from [>NO•]0 by less than 10%. It's important that hereafter consumption rates of injected >NO• and one being formed from hydroxylamine are almost the same (see curves 1 and 2 at Fig. 4).

As already been mentioned (see Section 3), if chains propagates by HO2•, >C(OH)O2 •, or >CH–CH(OO•)N< radicals then inhibition proceeds as disproportionation of these radicals with >NO•. Herewith hydroxylamine being injected quickly transforms to >NO• which almost doesn't consume but effectively inhibits an oxidation of corresponding substrates. So in this case the inhibiting factor values are more than 100 (Kovtun et al., 1974). Therefore it should be considered that in case of styrene's oxidation the way of irreversible >NO• consumption is reaction (4).

276 Nitroxides – Theory, Experiment and Applications

[>NO•

1,0

]/[>NO•

]0

6, 2 – 5.9∙10–6, 3 – 1.2∙10–5, 4 – 2.8∙10–5 (Pliss et al., 2012).

0,0

0,5

the same (see curves 1 and 2 at Fig. 4).

time according to the following equation (Loshadkin et al., 2002):

1 / ln

M–1∙s–1 (Pliss et al., 2012), was later used in mechanism's kinetic modeling.

**Figure 2.** Kinetics of >NO• (III) consumption during styrene oxidation (in relative coordinates); points: experimental data, curves: a result of simulation. *P*o2 = 20 kPa; *W*i = 1.0∙10–8 M∙s–1; [>NO•]0, M: 1 – 2.3∙10–

0 100 200 300

Time, min

Styrene's oxidation can be effectively inhibited by hydroxylamines. This assertion confirms with distinct induction period on oxygen consumption's kinetic curve (typical one represented at Figure 3 (Pliss et al., 2012)). Rate constants of such antioxidants' reactions with peroxide radicals can be determined from inhibited oxidation's rate dependence on

> 0 0 5.2 0 0 2

(5)

1 2

3

4

•, or

1 / [M] *WW W kW F t const WW W k*

But the problem is that oxidation rate remains significantly lower than *W*0 after the inhibiting period (see Figure 3) since >NO• being formed is an inhibitor of oxidation itself. However the kinetic modeling shows that in this case in equation (5) we can substitute *W*<sup>0</sup> value to the value of oxidation's rate at after-induction period. Herewith *k*5.2 determination error is less than 10%. Calculated *k*5.2 mean value at 323 K for >NOH (III), equal to 4∙106

Dependences of >NO• consumption and its accumulation from corresponding >NOH on time during styrene's oxidation at *P*o2 = 20 kPa presented at Fig. 4 (ESR-spectroscopy method). It's seen that hydroxylamine being injected quickly transforms to >NO• and its maximum concentration differs from [>NO•]0 by less than 10%. It's important that hereafter consumption rates of injected >NO• and one being formed from hydroxylamine are almost

As already been mentioned (see Section 3), if chains propagates by HO2•, >C(OH)O2

>CH–CH(OO•)N< radicals then inhibition proceeds as disproportionation of these radicals

**Figure 3.** Kinetics of O2 consumption during styrene's oxidation: 1 – without inhibitor; 2 – [>NOH (III)] = 2∙10–5 M; 3 – anamorphous of curve 2 in coordinates of equation (5); *P*o2 = 20 kPa; *W*i = 1.0∙10–8 M∙s–1 (Pliss et al., 2012)

**Figure 4.** Kinetics of >NO• (III) consumption (1) and its accumulation from corresponding hydroxylamine (2) during styrene's oxidation: dots – experimental data, curves – modeling results; [>NO• (III)]0 = [>NOH (III)]0 = 2.9∙10–5 M; *P*o2 = 20 kPa; *W*i = 1∙10–8 M∙s–1 (Pliss et al., 2012)

As it was mentioned above (see section 3.1), reaction (4) may proceed both as recombination and as disproportionation. Target experiment's results presented at Figure 5. Rate of >NO• (III) consumption in styrene (atmosphere of argon) during initiated oxidation was equal to initiation rate, and this >NO• consumption was proceeded up to detection limit of ESR-spectrometer ( 1∙10–7 M). After that argon was substituted to oxygen and, as can be seen at Figure 5, >NO• signal was appeared again.

These cycles were repeatedly detected several times until we reached the spectrometer's detection limit. Each time >NO• was recrudesced in share about 25–30% of its initial concentration. This fact confirms the assumption that reaction (4) also may proceed as disproportionation due to β-C–H bond of styrene's alkyl radical (~CH2C•HC6H5) with olefin's formation ~CH=CHC6H5 (M–H):

$$\text{M}^\* + \text{>NO}^\* \longrightarrow \text{>NOR} \tag{4.1}$$

$$\rm M^{\bullet} + \rm \rm NO^{\bullet} \longrightarrow \rm M\cdot\rm \rm \tag{4.2}$$

Thus >NO• signal may appear due to the reaction (5.2) upon oxygen blow (see Scheme 5 below).

**Figure 5.** Kinetics of >NO• (III) consumption in styrene: [>NO• (III)]0 = 2.8∙10–5 M; *W*i = 1∙10–8 M∙s–1 (Pliss et al., 2012)

Another one probable reason of this effect (Figure 5) is reaction MO2• + >NOR product + >NO• (5.3) that was first proposed in (Denisov, 1982). As >NO• regeneration source, this reaction was studied in detail for reactions of some alkoxyamines >NO• (I) with peroxide radicals of cumene and cyclohexylmethyl ether at 338 K (Kovtun et al., 1974). Obtained *k*5.3 values (1 – 26 M–1∙s–1) suggest that >NOR being studied in (Kovtun et al., 1974) are weak inhibitors. This conclusion also is confirmed with the dependences of styrene's oxidation rates on >NOR (I) and >NO• (III) concentrations (Pliss et al., 2012).

All of the experimental data and our previous results (Pliss et al., 2010a, 2010b, 2012) allow us to provide the following formal kinetic scheme of styrene's oxidation inhibited by aliphatic stable nitroxide radicals:

278 Nitroxides – Theory, Experiment and Applications

seen at Figure 5, >NO• signal was appeared again.

olefin's formation ~CH=CHC6H5 (M–H):

[>NO•

0

1

2

3

]10<sup>5</sup>

Ar

, mol/L

below).

et al., 2012)

As it was mentioned above (see section 3.1), reaction (4) may proceed both as recombination and as disproportionation. Target experiment's results presented at Figure 5. Rate of >NO• (III) consumption in styrene (atmosphere of argon) during initiated oxidation was equal to initiation rate, and this >NO• consumption was proceeded up to detection limit of ESR-spectrometer ( 1∙10–7 M). After that argon was substituted to oxygen and, as can be

These cycles were repeatedly detected several times until we reached the spectrometer's detection limit. Each time >NO• was recrudesced in share about 25–30% of its initial concentration. This fact confirms the assumption that reaction (4) also may proceed as disproportionation due to β-C–H bond of styrene's alkyl radical (~CH2C•HC6H5) with

M• + >NO• >NOR (4.1)

 M• + >NO• M–H + >NOH (4.2) Thus >NO• signal may appear due to the reaction (5.2) upon oxygen blow (see Scheme 5

**Figure 5.** Kinetics of >NO• (III) consumption in styrene: [>NO• (III)]0 = 2.8∙10–5 M; *W*i = 1∙10–8 M∙s–1 (Pliss

0 20 40

Time, min

Ar

Ar

<sup>O</sup> Ar O2 <sup>2</sup> O2

Another one probable reason of this effect (Figure 5) is reaction MO2• + >NOR product + >NO• (5.3) that was first proposed in (Denisov, 1982). As >NO• regeneration source, this reaction was studied in detail for reactions of some alkoxyamines >NO• (I) with peroxide radicals of cumene and cyclohexylmethyl ether at 338 K (Kovtun et al., 1974). Obtained *k*5.3 values (1 – 26 M–1∙s–1) suggest that >NOR being studied in (Kovtun et al., 1974) are weak inhibitors. This conclusion also is confirmed with the dependences of styrene's oxidation rates on >NOR (I) and >NO• (III) concentrations (Pliss et al., 2012).


**Scheme 5.** Detailed mechanism of vinyl monomers oxidation inhibited by nitroxide radicals

We've used this scheme for kinetic modeling (Pliss et al., 2012). Values of *k*1 – *k*4.1, *k*4.3, *k*5.3 (M– <sup>1</sup>∙s–1) were taken from the literature and values of *k*4.2 and *k*5.1 were obtained from modeling. Figures 2 and 4 shows that calculated curves are of satisfactorily consistent with experimental data. This indirectly confirms the reliability of Scheme 5.

Kinetic analysis shows that [M•] << [MO2•] when [O2] ~ 1∙10–2 M. In this case reactions (3.1), (3.2), (4.2), (4.3), and (5.3) can be neglected. Then the scheme including reactions (i), (1), (2), (3.3), (4.1), (5.1), and (5.2) can be described by equation:

$$\mathcal{W}\_i \left( \frac{\mathcal{W}\_0}{\mathcal{W}} - \frac{\mathcal{W}}{\mathcal{W}\_0} \right) = \frac{k\_4 \text{[} \text{>NO}^\bullet\text{]}\_0 \mathcal{W}\_0}{k\_1 \text{[}\text{O}\_2\text{]}} + \frac{2k\_5 \text{[} \text{>NO}^\bullet\text{]}\_0 \mathcal{W}\_i^{0.5}}{k\_{3.3}^{0.5}} \text{ > } $$

where *k*5 = (*k*5.1[>NO•] + *k*5.2[>NOH])/2[>NO•]0. If [>NO•] < 10–4 M then chain termination by reaction (4) can be neglected, therefore that scheme can be described by equation:

$$\frac{\mathcal{W}\_0}{\mathcal{W}} - \frac{\mathcal{W}}{\mathcal{W}\_0} = \frac{2k\_5 \text{[} \text{>NO}^\bullet\text{]}\_0}{\left(\mathcal{W}\_i k\_{3.3}\right)^{0.5}} \,. \tag{6}$$

Previously (Pliss et al., 2010a, 2010b) we've analyzed a simplified scheme including reaction (i), (1), (2), (3.3), (5.1), and (5.2). We've calculated the values of *k*5 = (5 ± 3)∙104 M–1∙s–1 for >NO• (I) – (V) in oxidizing vinyl monomers at 323 K and *P*o2 = 1∙105 Pa. These values are close enough to estimated in this present work value *k*5.1 = 2.5∙104 M–1∙s–1.

A fundamental question about the detailed mechanism of the reaction (5.1) remains open. By analogy with the oxidation of 1,2-substituted ethylenes and 1,4-substituted butadienes (Mogilevich & Pliss, 1990) we can assume that hydroxylamine's formation is facilitated by conjugation of β-C–H bond with peroxide bridge of styrene's polyperoxide radical:

$$\star\text{-CO-CH=CH{}(CHs)-CO}\star + \text{NO}^{\bullet} \longrightarrow \star\text{CO-CH=CH{}\bullet\text{-C}\text{H}\cdot + \text{NO}\text{H}\cdot + \text{O}^{\bullet}$$

Peroxide bridge is an important structural unit of ~MO2• radical. It alters the reaction center's electronic characteristics and increases the electrostatic term's contribution to the transition state's total energy (Denisov, 1996; Denisov & Afanas'ev, 2005; Mogilevich & Pliss, 1990). Probable reason of this effect is the difference in the triplet repulsion, which is close to zero in transition state of disproportionation reaction of MO2• with >NO• and is sufficiently large for the reaction of >NO• with nonconjugated C–H bond of hydrocarbon (Denisov, 1996). The latter probably explains the fact that aliphatic nitroxide radicals inhibit the hydrocarbon's oxidation via reaction with alkyl radicals only.

The results obtained in the present study draws attention to the results gained for biological systems where it is assumed that reaction of aliphatic >NO• with peroxide radicals proceeds via >NOOOR adduct formation decomposing to corresponding oxoammonium cations (Barton et al., 1998; Goldstein & Samuni, 2007; Offer & Samuni, 2002). The probability of such intermediate's existence is also considered in quantum-chemical analysis (Hodgson & Coote, 2010; Stipa, 2001). Further regeneration of nitroxide radicals may be due to reaction of oxoammonium cations with common biological reducing agents (Goldstein & Samuni, 2007; Offer & Samuni, 2002).

Direct reaction MO2• + >NO• MOOON< that results to stable trioxide's formation is seems quite doubtful for aliphatic >NO• in organic phase at moderate temperatures ( 373 K). First, it's easy to reject on the base of kinetic reasons cause in this case the kinetics of >NO• consumption and stoichiometry of chain termination would have a different nature than those observed in numerous studies (Browlie & Ingold, 1967; Kovtun et al., 1974; Pliss et al., 2010a, 2010b, 2012; Pliss & Aleksandrov, 1977). Second, it's easy to refute by direct quantum-chemical calculations (DFT B3LYP/6-31G\*, Table 7). It's easy to see that peroxide radicals' addition to >NO• is thermodynamically unfavorable.



### **5. Conclusions**

Thus, we must conclude that reaction of nitroxide with peroxide radicals plays an important role during styrene's oxidation in presence of aliphatic stable >NO•. This reaction proceeds probably as disproportionation and results to a partial >NO• regeneration.

At the same time we emphasize that detailed mechanism of chemical and biological oxidation processes inhibited by stable nitroxide radicals is still far from being established. Therefore kinetic experiments on the key reactions involving nitroxide radicals and its conversion products (hydroxylamines, alkoxyamines, oxoammonium cations) in solutions of organic substrates and in biological systems must be carried out to solve this problem.
