CH^)^ CN = NCH(CH3)2 and (CH3)*CH N ( c H ~ ) c

0 downloads 0 Views 5MB Size Report
The thermal decompositions of alkenes are very complex in consequence of ... produced by addition of a small radical (A) to the N=N double bond: x + YN=NY + YN(X)NY. (a) is consumed not only in H-abstraction, combination and disproportionation reactions, but also by breaking of a C-N bond different fiom the one just ...
J. Chim. Phys. (1999) 96, 591-609 O EDP Sciences. Les Ulis

A kinetic study of the reaction between hi3radicals and azoisopropane; reactions of the radicals en3, CH^)^ CN = NCH(CH3)2and (CH3)*CH~ N ( c H ~ ) c H ( c H ~ ) ~ Z. ~ i r a l ~M: ' , ~ o r ~ é n y iand ' ' L. seres2

' lnstitute of Physical Chemistry, University of Szeged,

P. O. Box 105, 6701 Szeged, Hungary Department of Chemistry, J.Gy. Jeachers Training College, P.O. Box 396, 6701 Szeged, Hungary (Received 22 May 1998; accepted 18 January 1999)

Correspondence and reprints.

Nous avons étudié la réaction des radicaux CH^ générés à partir du péroxyde de ditertiobutyle avec I'azoisopropane (AIP) par analyse des produits. Nous avons détenniné la régioselectivité des radicaux kH3 avec les centres radicalaires du carbone et d'azote de (cH~)~CN=NCH(CH~)~:

CH^ + (CH& CN=NCH(CH~)~ +

(CH3)3CN=NCH(CH3)2

+ (CH3)2C=NN(CH3)CH(CH3)2

klslk19 = 1.95 + 0.24

(18) (19)

Nous avons déterminé les paramètres d'Arrhénius des réactions d'abstraction d'hydrogène et de décomposition suivantes:

e

~3

+ AIP + CH4 i- (CH& CN=NCH(CH~)~

(cH~)~cHNN(cH~)cH(cH~)~ -+ 2 - C 3 ~ ,+ (CH3)2CHN=NCH3 par rapport à ceux des réactions de recombinaison suivantes:

CH^

2 + C2& 2 (cH~)~cHNN(cH~)cH(cH~)~ +

(3) (23) (13)

(CH3)2C~(CH~)N(CH(CH3)2)N(CH(CH,),)N(CH (26) Les expressions obtenues sont:

Z. Kirdly et al.

592

où 8 = RT ln 10 et 4 est la proportion de recombinaison croisée CH3 et (cH~)~cH~(cH~)cH(cH~)~Mots-clés: Reaction en phase gazeuse, méthyle radical, abstraction d'hydrogène, décomposition, azoisopropane, péroxyde de ditertiobutyle, réaction initiée.

ABSTRACT The reaction of e H 3 generated from di-tert-butyl peroxide with azoisopropane (AIP) in the temperature range 395-450 K was investigated by product analysis. The regioselectivity of CH3 with the carbon and nitrogen radical centres of (CH3)2~ N = N C H ( C Hwas ~ ) ~ determined: CH3+ CH^)^ ~ N = N C H ( C H ~ ) ~ (CH3)3CN=NCH(CH3)2 (1 8) (CH3)2C=NN(CH3)CH(CH3b (19) klslk19= 1.95 0.24 The Arrhenius parameters of the H-abstraction and decornposition reactions

*

CH^

+ AIP + CH4 + (cH~)~CN=NCH(CH~)~ (3) (cH,)~cHNN(cH~)cH(cH~)~ + 2- C 3 ~ +7 (CH3)2CHN=NCH3 (23) relative to the combinations 2 CH3 + C2& (13) 2 ( c H ~ ) ~ c H ~ ( c H ~ ) c H ( c H ~+ )~ (CH3)2C~(CH3)N(CH(CH,),)N(CH(CW9)2)N(CH)() (26) were determined:

where B = RT In 10 and

4

is the cross-combination ratio of e H 3 and

(cH~)~cHT;JN(cH~)cH(cH~)~. Key-words: Gas-phase reaction, methyl radical, H-abstraction, decomposition, azoisopropane, di-tert-butyl peroxide, initiated reaction. INTRODUCTION The thermal decompositions of alkenes are very complex in consequence of the oligomerization, intramolecular isomerization, decomposition and cyclization J. Chim. Phys

Reaction between methyl radicals and azoisopropane

593

processes that are frequently encountered. The complexity of such reactions can be reduced if the temperature is decreased and initiators are used [l-61, or if Hz is added, when hydrogenolysis-hydrogenization is more important than the decomposition [7,8]. (E)-azoalkanes are fiequently used initiators in both photolytic and thermal radical reactions [ 5 ] . Except when they are used as additives in very small amounts, the alkyl radicals produced react with the parent azoalkanes by Habstraction and addition to yield N-containing radicals, and this modifies the composition of the products [9,10]. Under the conditions where azoalkanes undergo thermal decomposition, some of the products may be unstable and this, if unrecognized, makes the evaluated data unreliable. If the reaction of an azoalkane is initiated with another type of initiator, e.g. ditevt-butyl peroxide (PODBT), which decomposes at a temperature some 100 K lower, most of the products are stable and the roles of elementary reactions that may also be important at higher temperatures can be revealed. An interesting feature of the initiated decompositions is that the radical produced by addition of a small radical (A) to the N=N double bond:

x + YN=NY + YN(X)NY

(a)

is consumed not only in H-abstraction, combination and disproportionation reactions, but also by breaking of a C-N bond different fiom the one just formed, i.e.

YN(x)I;IY

+ m=NY4- Y

(b)

The bulkier the group Y, the greater the possibility that the bond N-Y will break. This is the reason why such a reaction was not significant in the PODBT-initiated decomposition of azoethane below 440 K (61. In the present paper, we report the results of a study in which the.PODBTinitiated thermal reaction of (E)-azoisopropane (AIP) was investigated in a

594

Z . Kiraly et al.

temperature range where AIP itself does not behave as an initiator. Besides an elucidation of the roles of some elementary reactions that may also be important at higher temperatures, relative rate constants of various elementary reactions were determined.

EXPERIMENTAL AIP was synthesized by known methods [Il], and was purified by preparative gas chromatography. Its purity was better than 99.6%. PODBT was purchased from Sigma-Aldrich (FLUKA) and was purified by standard methods prior to use. Its purity was better than 99.8%. The experimental and analytical methods were similar to those described earlier

[6].The reaction was studied in a static system in a quartz reaction vessel at the initial concentrations [AIPjo = 1 . 2 1~

mol dm-3and [PODBVo = 1 . 3 5 1~o4 mol

dm" and in the temperature range 395-450 K. In order to ensure perfect mixing, the reactants were stored at 323 K in a mixing vessel for 1 h prior to admission into the reaction vessel. Two samples were taken for analysis. The reaction mixture was expanded through a heated sample line into a gas sampling bulb (eu. 100 cm3) containing n-butane as intemal standard, and the rest of the mixture was fiozen into a solution of n-octane as intemal standard in isooctane. In most of the experiments, the conversion for AIP was 1-2%, and that for PODBT 10-15%. The products were identified by gas chromatography/mass spectrometry (GCIMS) (VG Micromass, ZAB-SEQ MS EI/CI SRC) and with a Bmker AM-400 nuclear magnetic resonance spectrometer. Quantitative analysis of products was performed with a Hewlett-Packard 5890 Series II GC equipped with a Barne ionization detector (FID). Separation of the Ci-Cs fraction was performed with a 3

m ( @ 3.2 mm) stainless-steel column filled with Durapack-n-octanelPorasil-C. Separation of the higher products was carried out on the liquid samples with a

Reaction between methyl radicals and azoisopropane

595

50 m (@ 0.32 mm) HP1 fiised silica capillary column (Hewlett-Packard) with

temperature programming. The product composition as a function of temperature is shown in Figs. 1-4. The points in the plots were obtained from the expression X;- = 100 cilC ci,where Xi and ci

were the yield and the concentration of the i-th product at the longest reaction

time at each temperature. These plots suggests that ( c H ~ ) ~ ~ N = N c H ( c H ~(R) ) ~ (Scheme 1) is a radical of major importance in this system. A self-combination product of a similar azo radical ( C H ~ ~ H N = N C H ~ C was H~)

observed earlier [6].Of the three possible self-combination products of R, analysis indicated only one, identified as the compound formed in the combination Rc-&

[GC Kovats retention index 903.8 (50 m FSOT HP-1, 60 OC)]. (Rcand RN denote the structures produced when the radical R undergoes combinations at the carbon and nitrogen radical centres, respectively.)

Figure 1 : Product composition of the reaction at direrent temperatures obtained with [PODBTIo = 1.35x1U4 mol clin" and [AIPIo = l.2xlU3 mol dm". MRç = (CH3)3CN=NCH(CH$a MRN = (Ci-fj)2C=NN(CH3)CH(CH3)2. = conversion of PODBT. J. Chim. Phys.

Z. Kiraly et al.

0.5

1

@

!

1

0.0 ' 400

I

-

6

v

? 410

9

1

A

A

420

=iHi

+

A

R'P.1

A DMBA 450

440

430

TIK

Figure 2 : Product composifion of the reaction a f dryerent temperatures under the same concentrafionconditions as in Fig. 1. RNH-, = (CH3)2C=NN=C(CH3)2. R 'P., = (CH3)2CHN=NCH3, DMBA = (CH3)2CHCH(CH3)r, = conversion ofAIP.

-7-

O2iC*

0.1 400

410

.

430

420

I R ' H a

6

440

.

.

1

I 450

TIK

Figure 3 : Prodttct compositiorl of the reaction at differenf temperatures under the same concentration corzditions as in Fig. 1. R 'H = (CH3)2CHNHN(CH3)CH{CH3)2, PRc = (CH3)2CHC(CH3)2N=NCH(CI13)2. PRN = (CHJ2C=NN[CH(CH3)2]2. J. Chim. Phys

Reaction between rnethyl radicals and azoisopropane

597

Figirre 4 : Prodrct composition of the reaction ut diyerent temperattrres under the same concentration conditions as in Fig. 1. RNH = (CHd2C=NNHCH(CH3)a, RrRc =(CHJ~CHN=NC(CHJ)~C(CH~)~N=NCH(CHL)~ MR ' = ( c H j ) ~ C f f N ( c H N(CHdCH(CHd2. ~)

(E)-(CH3)2CHN=NCH3, a product of ( C H ~ ) ~ C H ~ ( C H ~ ) C H ( C (HR, ) ~ decomposition, was synthesized and its formation in the reaction mixture was confirmed by GC and GC/MS. Difficulties with the syntheses of some azo, hydrazono and hydrazino compounds led us to use the effective carbon number (ECN) concept [12] to convert the FID signal to concentration. The influence of the groups -N=N-, >C=N-and

>N-N< on the ECN was evaluated from the response factors of these compounds [13,14]. A possible pressure dependence in the rate of formation of the products was

investigated by the addition of up to 40 kPa COz to the reaction mixture. Merely minor effects were observed on the rates of formation of acetone and tert-butanol. For al1 other products, any possible effect was masked by experimental errors.

2-4 experiments were carried out with the same initial concentrations at each

Z. Kirhly et al.

598

reaction temperature. In the early stages of the reaction, the concentration vs. time plots of the key compounds were virtually linear, and the rate was determined as the ratio of product concentration to reaction time, i.e. r = dc/dt z AclAt. The concentration vs. time plots of different products at 423 K are shown in Figs. 5-8. The Arrhenius parameters were evaluated from the exponential form of the equation. The errors quoted are standard deviations.

Scheme 1. Mechanism of the reaction of PODBT and AIPa Initiation: PODBT + 2 (cH~)~cC> ((-sui>) t-BUO + CH3 + (CH3)2C0 H-abstraction: è ~ +3AIP -+ C& + ( c H ~ ) ~ ~ N = N c H ( c H(R) ~)~ (cH~)~c6 + AIP + (CH3)sCOH (t-BuOH) + R R' + AIP -+ (CH3)2CHNHN(CH3)CH(CH3)2 (R'H) + R Addition: CH^ + AIP + ( c H ~ ) ~ c H ~ ~ ( c H ~ ) c H ( c (R') H,~ Disproportionation: e ~+3 R + Cfi + (CH3)2C=NN=C(CH3)2 (RNH-1) 2 - è 3 ~ 7+ R + C3H8+ RNH-, e ~+3 R' -+ C& + (CH3)2C=NN(CH3)CH(CH3)2 (MR,) 2 - C 3 ~+7 R' + C3H8+ MRN 2 2- C 3 ~ 7-+ C3Hs+ C3H6 ~ H ~ + ~ -+ C ~Cfi+C3i& H ~ Combination: 2 CH^ + C2& 2 2- C3H7 + (CH3)2CHCH(CH3)2 CH3 + 2- C 3 ~ 7 + (CH3)3CH 2 R + (CH~~CHN=NC(CH,)ZC(CH~)~N=NCH(CH~)~ (RcRc) + [(CH3)2CHN=NC(CH3)2N(CH(CH3)2)N=C(CH3)21 (&RN) + RNHeIf N2 + 2 2- k 3 ~ 7 ~3 + R -+ (CH3)3CN=NCH(CH3)2 (M&)

(1)

(2) (3) (4) (5)

(6)

(7) (8) (9) (10)

(1 1) (12) (13) (14)

(15) (16) (17) (18)

J. Chim. Phys.

Readion beîween methyl radicals and azoisopropane

+ 2- C

3 ~+7R

MRN

-+ (CH3)2CHC(CH3)2N=NCH(CH3)2 (Pe) + ( C H ~ ) Z C = ~ [ C H ( C H ~ (PRN) )Z]~

k~~ + Rr + (CH3)2CHN(CH3)N(CH3)CH(CH3)2 (MW) Decomposition:

R' Re& Isomerization: AIP

+ +

2-C 3 ~ +7(CH3)2CHN=NCH3 (RP-1) 2 2 - C 3 ~+, 2 N2 + (CH3)2C=C(CH3)2

+

(CH3)~C=NNHCH(CH3)2 (RNH) 1

" Abbreviations are based on: P = 2 - C 3 ~ , . &= (CH3)*CN=NCH(CH3)2,

Figure 5 :Products ofthe reaction of PODBTand AIP at 423 K obtained with [PODBTjo = 1 . 3 5 ~ 1 ~mol ~ 'dmJ and [AIPIo = 1 . 2 ~fT3 1 mol dmJ. MRc = (CHj)jCN=NCH(CH)a

AA

Z. Kiraly et al.

MR'

-8

.---A

O

5

10

15

20

25

30

A

35

40

45

50

55

Figure 6 :Products of the reaction of PODBT and AIP at 423 K obtained under the same concentration conditions as in Fig. 5. PRN = (CH3)2C=MV[CH(CH$2]2, MR ' = (CH3)2CHN(CHdN(CH3)CH(CH3)z,DMBA = (CH3)2CHCH(CH2)2, PRc = (CH3)2CHC(CH3)2N=NCH(CH2)t.

Figure 7 :Products of the reaction of PODBT and AIP at 423 K obtained under the same concentration conditions as in Fig. 5. RNH = (CH3)2C=NNHCH(CH2)2, R 'P.,= (CH3)2CHN=NCH3.

Reaction between rnethyl radicals and azoisopropane

Figure 8 :Products of the reaction of PODBT and AIP ut 423 K obtained under the same concentration conditions as in Fig. 5. MRN = (CH3)2C=NN(CH3)CH(CH3)2, RNH-] = (CH3)2C=NN=C(CH3)2, R 'H = (CH3)2CHNN(CH$CH(CH$2, RcRc = (CH3)2cW=NC(CH3)$(CH3)2N=NCH(CH3)~.

RESULTS AND DISCUSSION Mechanism of the reaction The formation of the most important products can be explained by the mechanism presented in Scheme 1. (The abbreviations used are also included.) It follows fiom the Arrhenius parameters of initiation by AIP [5j that such a process yields radicals at a rate some two orders of magnitude lower than the rate for

PODBT under the present experimental conditions, and the respective step has not been included in the mechanism. The radical

t- BU^ produced in the initiation step yields (CH3)2C0,k~~ and t-

BuOH on decomposition and H-abstraction, respectively. Only reactions (2) and (4) of t-BUOhave been included in the mechanism, because the ratio

0.98

* 0.14, where ro3

respectively:

and r , - are ~ ~the~ rates of formation of

r,-~ /r,-BuO ~ is

CH^ and t-BUO,

Z.Kiraly et al.

602

At 423 K, the

formed in reaction (2) is mostiy converted into methane

(64%), MRÇ and MRN,the combination products of

CH^ and R (17%), and ethane

(1 0%) at 423 K.

Within experimental error, r c ~/r,-BuO 3 was temperature-independent. These facts demonstrate that (a)most of the t- BU^ was converted to

6 ~ (i.e. 3

PODBT is a relatively 'clean' source of CH3), and (b) almost al1 of the

e~~

formed was incorporated in the products measured (i.e.no important loss of

CH^

in unrecognized side-reactions occurs). Addition of CH3 to >C=C< double bonds is a major reaction under similar experimental conditions, but semiempirical quantum-chemical calculations indicated much higher enthalpies of activation for

CH^

additions to -N=N-

double bonds [15]. Thus, formation of products for the adduct radical R' in small amounts is in agreement with the expectations. The combination products allowed us to eliminate radical concentrations fiom the rate equations

~ C ~ =H k13[CH312, , r

m = ~ ~ ~ [ C H ~ I [ R ' ] ,= ~ I ~ [ C H ~ ] [ R ] .

In this way, various relative rate constant ratios could be determined.

Reactions of R Radical R formed from AIP in different H-abstractions was converted into M& and MRN, the combination products with

CH,

(43%), the dimerization product

RcRc,(35%), and PRÇ and PRN,the combination products with 2-t3H7(5%). Thus,

and R yield important products in self- and cross-combinations. J. Chim. Phys.

Reaction between rnethyl radicals and azoisopropane

603

These numbers and the formation of different combination products suggest that the disproportionation reactions between

k~~ and R may also be important in

the removaf of R. One of the possible disproportionation products, RNH-,, was identified, but the other disproportionation product, &Hm,, was not observed, probably because it elutes together with the broad peak of AIP. R can react in three different ways via self-combination to yield K R c , &RN

and RNRN.Formation of the combination product RcRcwas observed in this work. As 2,3-dimethyl-2-butene [a decomposition product of Rc&,

reaction (24)] was

formed in small amounts (c I W A ] ) , the self-combination product && identified in this study was relatively stable. There is evidence of the occurrence of both combinations C-C and C-N of diazaallyl radicals [16], but only the selfcombination product C-C was observed fkom the radical analogous to R in the PODBT-initiated reaction of azoethane (AE) [6]. In the self-combination of tertc ~ H ~ N = N ~ ( c Hthe ~ )preferred ~, C-C combination was explained by steric effects due to the bulky tert-CsH9 group [16]. The absence of any self-combination product RcRN in the present study can be explained either by a negligibly small rate of formation due to steric effects, or by the low thermal stability of the product. Decompositions of &Rc, RcRN and R' yield 2 - e 3 ~ incorporated , in products C36, C3&, P h , PRN, etc. As the products incorporating 2 - è 3 ~ 7are apparently primary, reactions (23) and (24) seem unlikely to give a full explanation of the formation of these species. We assume that the combination C-N of R, followed by fast decomposition of the product. RcRNmayserve as a hrther source of 2 - t 3 ~ [reaction 7 (1 7)]. The energy of activation for the C-N rupture is (estimated from Ea(AIP)

=

Ca.

201.9

50.2 kJ mol-' lower than that for AIP Id mol-'

C4H9N=NC(CH3)2]2) = 147.4 kJ mol-' [16]).

[4] and AHd3:

([tert-

2. Kiraly et al,

604

On decomposition, the combination product &RN yields RNH-,,an important product in the reaction. RNH.,is also formed in reaction (17).

On substitution for the radical concentrations and rearrangement, we have

From the rates of accumulation of different products, the parameters of the latter equation were evaluated by the use of multilinear least squares:

k7/kI8= 0.138 f 0.033 and kl3kI7/k1g2 = 0.086 $: 0.039. It follows from these parameters that reaction (7) makes a 62% contribution to RNH-iformation at 423 K.

Relative reactivities of

? and 3 radical centres of R in combinations

The combination reactions (1 8) and (19) are expected to be major sources of

M& and MRN. Further sources of MRN are the disproportionations of R' in l

reactions (9) and (10). The contributions of these processes were expected to be small and they have been neglected. Then:

From rmc and r m of

CH:,with the

and

determined at different temperatures, the regioselectivity

fi radical centres of R was determined:

J. Chim. Phys.

Reaction between methyl radicals and azoisopropane

605

The combination CH3-& is therefore preferred over CH3-RN,in agreement with earlier results [6].

H-abstraction from AIP by

CH^

The rate of formation of methane in reaction (3)is

where Y , = 0 . 1 3 8 ,~rl2~=~kizlkist i - c 4 H 1and 0 kI2/kl5 = 0.163 [17]. The main source of C& is reaction (3); at 423 K the contribution of reaction (7) is 2.5%, and that of reaction (12)is 0.2%. The contribution of reaction (9) was expected to be small and it was neglected. Substitution for [ k ~ 3as] above yields:

The temperature dependence of the relative rate constants was evaluated by the use of weighted nonlinear least squares [w = 1/J log[(k3/

(y = t 3/ k:?)]

(Fig. 9):

kf3)/ m 0 1 ' ~ ' ~ d r n ~ ' ~ s =~ "(3.1 ~ ) ] f 0.3) - (30.6I2.4)kT mol-'/ 8

where 8 = RT In 10,and the error limits were increased frorn the calculated ones to allow for the unrecognized systematic errors. There are no published data on reaction (3) in the literature. However, the data reported for analogous reactions of azomethane (AM)and AE compare well with those obtained here (Table 1). The only exceptions are the Arrhenius parameters determined by Durban and Marshall [19] on AM, although the rate constant ratio calculated for 420 K is also in good agreement with the other data in this case.

Z.Kiraly et al.

606

Table 1: Kinetic data for H-abstraction reactions from azoalkanes by

EH,.

KIT Figure 9 :Arrhenius plot of H-abstractionfrornAIP by

~ 3 .

Decomposition of R '

The addition of

è~~to AIP is practically irreversible in this temperature range.

In the dissociation of R', the bond (CH3)*C-N< [reaction (23)] rather than the bond CH3-Nc is expected to break, and even the latter reaction is a minor one: of al1 the products of R', R'P-I accounts for only 14.5% at 423 K. The rate coefficient k23 relative to the rate constant of combination of R' could

Reaction between methyl radicals and azoisopropane be determined from the kinetic equation

where

[wwas eliminated as above. On rearrangement, we have

On substitution for k!j/k22 via the cross-combination ratio for

CH^ and R':

where kz6 is the rate constant of combination of R':

2 R' + (CH3)2CHN(CH3)N[CH(CH3)23N[CH(CH3)2IN(CH)

(26)

The expression on the right-hand side was evaluated and the temperature dependence was calculated by weighting [according to w = lly2, y

=

k23/(4k:52)]

(Fig. 10). 312 112

I ~ [ ~ ~j'lm01'~dm~ ( c ~ s'\ ~] = (8.8 1 0.5) - (1 10.0 16.5) kJ m o l . ' l ~

To Our knowledge, there are no data for reaction (23) in the literature. The only

daturn on a similar reaction, where 2-t3H7 rupture from an alkyl radical occurs, is

+~ )C3& + 2 - è 3 ~ ?=] 118 kJ mol-' [21]. Here, the E,[(CH~)~CHCH~~HCH energy of activation compares well with the present datum. J. C h h . Phys.

Z. Kiraly et al.

Figure IO :Arrhenius plot of decomposition of R '.

CONCLUSION Lowering the reaction temperature of the thermal reaction of AIP by use of

PODBT as an initiator resulted in the reduction of the complexity of the reaction, and formation of specific products was observed. The product composition suggests that R, the radical formed from AIP in H-abstraction reactions, was a dominant radical in the system. The C-C self-combination product of R was routinely measured, however, formation of the C-N and N-N

combination

products was not observed. Cross-combinations of

CH^ with the carbon and nitrogen radical centres of R

favour combination with the carbon radical centre. From the rates of formation of the products relative rate constants on some of the elementary reactions could be determined. J. Chim. Phys.

Reaction between methyl radicals and azoisopropane

ACKNOWLEDGEMENT This work was supported by a grant from the Hungarian Research Foundation (OTKA T0 16044). REFERENCES 1 Marshall R.M., Rahman L. (1977) In?. J .C'hem. Kinet. 9,705-724. 2 Scherzer K., Claus P., Karwath M. (1985) Z. Phys. Chem. 266,321-328. 3 Dobé S., Bérces T., Réti F., Marta F. (1987) Int. J. Chem. Kinet. 19,895-921. 4 Gorgényi M., Seres L., Fischer R., Scherzer K. (1995) J.C.S. F a r a d v Trans. 91,1303-1312. 5 Engel P.S. (1980) Chem. Ra). 80,99-114. 6 Gorgényi M., Seres L. (199 1) J. C.S. Faraday Trans. 87, 1 827- 1830. 7 Collongues C., Richard C., Martin R. (1983) In?. J. Chem. Kinet. 15,5-23. 8 Barbé P., Martin R., Perrin D., Scacchi G. (1996) Int. J. Chem. Kinet. 28, 849-863. 9 Martin G., MaccoH A. (1977) J.C.S. Perkin II. 1887-1893. 10 Strausz O.P., Berkley R.E.,Gunning H.E. (1969) Can. J. Chem. 47,3470-3474. 11 Renaud R., Leitch L.C. (1954) Can. J. Chem. 32,545-549. 12 Sternberg J.C., Gallaway W.S., Jones D.T.L. (1962) in Gas Chromatography, edited by Brenner N.C., Callen J.E. and Weiss M.D., Academic Press, p. 23 1. 13 Gorgényi M., Fekete Z., Seres L. (1989) Chromatographia 27,581-584. 14 Kiraly Z., Kortvélyesi T., Seres L., Gorgényi M. (1996) Chromatographia 42, 653-659. 15 Komélyesi T., Seres L. (1995) React. Kinet. Catal. Lett. 56, 371-376. 16 Engel P.S., Wang C., Chen Y., Rüchardt C., Beckhaus H.D. (1993) J. Am. Chem. Soc. 1 1 5,65-74. 17 Terry J.O., Futtrel J.H. (1967) Can. J. Chem. 45,2327-2333. 18 Arican H., Arthur N.L.(1 983) Aust. J. Chem. 36,2 185-2589. 19 Durban P.C., Marshall R.M. (1 980) Znt. J. Chem. Kinet. 12, 1031-1043. 20 Gorgényi M., Korîvélyesi T., Seres L. (1993) J.C.S. Faraday Trans. 89, 447-450. 2 1 Allara D.L., Shaw R. (1980) J. Phys. Chem. 9,523-559.

_

J. Chim. Phys.