Inhibition mechanism of nickel dialkyldithiocarbamates in oxidation of cumene

INHIBITION~..MECHANISYl OF NICKEL DIALKYLDITHIOCARBAMATES IN OXIDATION OF CUMENE* L. L. GERVITS, •. V. ZOLOTOVAand Yv,. T. D~.msov I n s t i t u t e of Chemical Physics, U.S.S.R. A c a d e m y of Sciences

(Received 21 May 1974) SALTS of metals (Zn, Ni, Cu) and dithiocarbamic acids are used to stabilize polymers and lubricants. Several papers have been published in recent years to explain the mechanism of their anti-oxidizing action. Carbamates react very rapidly with hydroperoxides by reducing them into alcohols [1-6]. There are indications that free radicals are formed in this reaction [6, 7]. However, important characteristics such as rate constant of the reaction of hydroperoxide with carbamate and the rate constant of radical formation by this reaction are so far unknown. Carbamates inhibit both initiation and photo-oxidation evidently by reacting with peroxide radicals [8-12]. The rate constant of the reaction of I~0~ with nickel carbamate has only been measured in one study [12]. For copper carbamate inhibiting ethylbenzene oxidation the reaction with alkyl radicals [10] is proposed. Thus, for earbamates there are no detailed quantitative methods available for evaluating those reactions, in which they take part in oxidation. Using nickel diethyl- and dibutyldithiocarbamates an attempt is made in this paper to give as complete as possible the kinetic characteristics of the action of carbamates on cumene oxidation, namely: studying quantitatively the interaction of earbamates with hydroperoxide, measuring the rate of formation of radicals by this reaction, obtaining quantitative characteristics of the reaction of carbamates with peroxide radicals. EXPERIMENTAL

The solvents used were purified by standard methods. Cumene hydroperoxide (ROOH)was distilled in vacuum under a pressure of 0.1 mmHg and at 52 ° in inert atmosphere. Nickel diethyldithiocarbamate (ENi) was recrystallized from chloroform and nickel dibutyldithioc~rbamate (BNi) from acetone. The reaction of cumene hydroperoxide with carbamate was carried out in ehlorobenzene in a thermostatically controlled glass reactor with a sampling * :Neftekhimiya 15, No. 1, 135-140, 1975. 27

L.L. GERVITS e$ al.

28

device with inert gas diffusion. The reaction was assessed by the concentration variation of carbamates over a period of time. Nickel carbamates used have typical absorption bands independent of the ligand. To determine carbamate concentration, the optical density (D) was m ~ u r e d at frequencies of 25,700 cm -1 with a molar extinction coefficient 0f 5.9× 103 l./m.cm and 30,700 cm-I with e : 3 . 4 × 104 1./m. cm. Cumyl hydropcroxide was analysed iodometrically. To measure the rate of radical formation by the reaction of ROOH with nickel carbamate in inert atmosphere, a free radical acccptor, N,N'-di-flnaphthylparaphenylenediamine (InH2) was added to the reaction mixture. This acceptor, reacting successively with two free radicals, was converted to clear quinonedi-imine (In), the concentration of which was measured spcctrophotometrically according to chlorobcnzene absorption in the solution with v----20,900 cm -1 (~:104 1./mole.cm). Experiments with different acceptor concentrations show t h a t Wxn~. is independent of [InH2], the concentration of which varied between 4 × 10 -~ and 9X10 -5 mole/1, with [ R O O H ] : I X I 0 -3 and [ E N i ] = 3 × 1 0 -5 mole/1. This D 1"4 1"2

1.

1.0

f'O

Z

i

0"8

0"8

0.6

0"6 : UOzhml

Z

5'

4l

0.4

fl'6 o.q

O.Z

0"2 9"2

j~~L

0

,

80

t

L i

,

,

j

ib'O

~

,

24[}

i

j

~

320 ~, nuTJ

:Fla. 1

.

_ 0

0 20

~0

GO

80 t, m m

FIO. 2

FIe, 1. Kinetics of oxygen absorption in auto-oxidation of cumene in the presence of nickel dibutyldithiocarbamate at 106°C. [BNi], mole/l: 1-- 0; 2-- 1.9 × 10-5; 3-- 3.6 x 10-s~ 4--6"3X 1 0 - 4 . :FIG. 2. Kinetic curves of consumption of nickel diethyldithiocarbamate by reaction with cumene hydroperoxide in benzene at 23°C. [ENi]0=2.5 × 10-4 mole/L; [ROOH]0× × 10-3 mole/1.; 1 -- 0.8; 2-- 1.6; 3-- 3-2; g-- 6.4; 5-- 8.0; 1"-5"--rectification of kinetic curves in semi-logarithmic coordinates.

Inhibition mechanism of nickel dialkyldithioearbamatcs

29

means that the acceptor in our experiments was mainly used up b y the reaction with free radicals. Kinetics of initiated oxidation of cumene with carbamates were studied b y oxygen absorption using a manometer at 60 ° with azodiisobutyronitrile (AIBN) as initiator. The stoichiometric coefficient of inhibition f was calculated from the inhibition period v ( f = Wiz/[Carbamate], from the intersection of t -l on the curve A [02]-1-t-1 and from the rate of consumption of carbamate Wo (f----Wi Wc-1). To calculate the rate constant of the reaction of earbamate with RO~ kinetic absorption curves of oxygen were rectified in coordinates

A [Os]-l-t-1. The gradient of this straight line A (A[Oz] -x) /ko[K]o A (t -1) -----kR~[RH]k~[AIBN]0

(I)

where kc and kRH are the rate constants of the reaction of peroxide radicals With carbamate and cumene, respectively, [K]0 is the initial concentration of carbamate. At 60 ° kR~----1.5 1./m.see [13], k ~ = l . 0 × 1 0 -5 see -1 [14]. RESULTS

Experiments on oxidation of cumene (3 ml) at 106 ° without initiator show that the kinetic curve of auto-oxidation is complex and depends on the initial carbamate additive (Fig. 1). First a period of intensive oxidation is observed, followed b y a smooth reduction of the rate of oxygen absorption to practically complete inhibition during a certain length of time and then repeated increase of oxygen absorption. With an increase of carbamgte concentration, the initial period of intensive oxidation decreases and the period of inhibition increases. Intensive oxygen absorption at the initial stages of oxidation may be due to the presence in cumene of traces of hydroperoxide and radical formation during the interaction of carbamates with hydroperoxide. To confirm this assumption, experiments were carried out into oxidation of cumene without an initiator and in the presence of cumyl hydroperoxide in a concentration of 2 × 1 0 -3 mole/1, and carbamate in a concentration of 2 × 10-4 mole/1, at a temperature of 110% The type of oxygen absorption curve coincides with the curves obtained in auto-oxidation, which is in agreement with the assumption concerning the initiating ability of the carbamated-hydroperoxide system. Experiments with free radical acceptors confirmed this. Reaction of carbamates with ROOH. Carbamates rapidly react with hydroperoxide even at room temperature. With a hydroperoxide excess kinetics of carbamate consumption are described b y a first order kinetic equation (Fig. 2). Rate constants calculated from the initial rate of consumption of nickel diethyldithiocarbamate and from semidogarithmic transformations, coincide within the range of experimental error. The rate of consumption of nickel diethyl-

110

L . L . GERVITSe$ al.

•

[ENi-]~lOSimole/L If

E'~

ho2,rnm

2,,.

z60

t,.I- I1I-

/ I

80

o

I

I

5

m

I#

I

I

I

I

i

~),iOZmin

,,

i

I

I

I

2o

,o

eo

eo

{ROlTi_l]_,,,O~ _nole/l. FIG. 3

/ .t It

I

I

t, m/n FIG. 4

FIG. 3. Dependence of the rate of radical formation on initial concentrations of cumy: hydroperoxide ( e ) and nickel diethyldithiocarbomate (O) in benzene in argon floWl e - - t e m p e r a t u r e 71.7°C; [ENi]~-6×10 -5 mole/L; O - - t e m p e r a t u r e 50°C; [ROOH]o

----1 × 10 -3 mole/1. FIG. 4. Kinetic curve in initiated oxidation of cumene in the presence of nickel diethyldithiocarbamate and rectification of the curve in coordinates A[O2)-l-At-1. Temperature 60°C; [AIBN]-----1.SX10-B mole/1, p0----760 mmHg: 1--[ENi]=0; 2--[ENi]----7×10 -I mole/L; 3--rectification of the kinetic curve. dithiocarbamate shows a linear dependence on the initial concentration of cumyl hydroperoxide. This indicates t h a t ENi and BNi react with ROOH b y a bimolecular reaction: Wc----k[ROOH][K ]. The rate constant is independent of the ratio [ROOH2]0/[K]0 on being changed from 2 to 250. As regards reactivity ENi and BNi are the same. The activation energy of this reaction, measured in the range of 20 to 85 ° is 5±0.7 kcal/mole, in a general form k----5×102 exp(--5000/RT)l./mole.sec. The stoichiometric coefficient of this reaction determined from the ratio between the hydroperoxide used up and the carbamate (ENi and BNi) is 5 . 7 i l . 7 and is independent of the initial component ratio which varied between 1 and 50; experiments were carried out at 23 °. The rate of radical formation in interaction of hydroperoxide with carbamates, measured from the rate of consumption of the free radical acceptor, is directly proportional to both corbamate concentration and hydroperoxide concentration (Fig. 3). Consequently, radicals are also formed by a second order reaction. Activation energy Ei~10.0-~-l.6 kcal/mole (temperature range between 25 and 72°). The rate constant of radical formation for ENi and BNi agree and in a general form i s

2 ~/Inll,

k~----[RO OI-I] [ENi]---- 3-2 × 10" exp(-- 10,000/RT)l.]m. sec

Inhibition mechanism of nickel dialkyldi~hiocarbamates

31

The probability of radical yield in the volume

e=kd/k=l.1

× ]03

exp (--5000/RT)

varies between 0.23 at 25 ° and 0.84 at 80 °, activation energy E, being evidently increased excessively b y 2-3 kcal/mole due to incorrect determination of E and E,. Absolute values of e and Ee are typical of radical yield in the volume from the "'cell" in hydrocarbons [14]. Results confirm that cumyl hydroperoxide interacts with ENi and BNi b y a homolytic mechanism and agree with conclusions of authors of other studies [5, 6] concerning t h e composition of products of interaction of several metal dialkyldithiocarbamates with cumyl hydroperoxide (temperature 60°; ethylbenzene, nitrogen). It is assumed [6] that hydroperoxide forms a complex with carbamate which breaks down to radicals. Authors have shown that the carbamate : hydroperoxide ratio in this complex is about 1 : 4. The mechanism of interaction of hydroperoxide with carbamates m a y be presented b y the system [1, 5]: ,

.,S\

/,S,

"~S/'

.S02--

"S ~//

/SO,,

/S\

,

.S.

S

S

S03

--SO-

' / S "sNI~\s//C" "/ " NR'~.+2ROOH--* R ~N"~/ I12NC~ t ~ S S\ N i /r \ S S\CNB'--a-2ROH ~ ,

,

It 2NC(

SOa\

/SO.

,

,

~Nir. ~CNR 2"--"R 2NC--S--CNR z + NiS04 ~ S / " S// il II S

S

R'2N--C--S--C--N R'o + ROOH --, R'~NC--S--S--C--N R'2 + ROI-I

]1

II

N

I!

S S S 0 R'=N--C--S--S--C--N R'2-~ ROOH--, R%NHS04-L other products S

0

The stoichiometric coefficient of the reaction determined b y us agrees with this system. Reaction of carbamates with RO~ and R" radicals. Nickel earbamates inhibit initiated oxidation of cumene (Fig. 4), which points to their participation in chain rupture. The rate constant kc/kRu was calculated from the ratio A (/1 [02]-1)/ /A (t-l) according to formula (1). The constant ratio is independent of the rate of initiation (1-2 X 10 -7 mole/1..sec), carbamate concentration (7-17 X 10 -4 mole/1.) and OfPo ' (0.2-1 arm). This indicates that the reaction of the peroxide radical with carbamate is the limiting stage of chain rupture. The ratio of kc/kRH=(l'7~0"3)× 10a at 60 ° for nickel diethyl- and dibutyldithiocarbamates, the absolute value being ko----(2-3-t-0-6)×10a 1./mole.see. The stoichiometric coefficient of inhibition f, measured b y various methods is 1 for nickel diethyl- and dibutyldithiocarbamates during oxidation with air and pure oxygen (Table). Results are in agreement with information in the literature. I t was shown [2] that zinc dibutyldithiocarbamate inhibits auto-oxi-

32

L . L . GmawTs et al.

dation of squalane (it follows from a comparative evaluation t h a t this carba' m a t e is weaker t h a n fl-naphthol). I t was indicated [10] t h a t Cu (II) hexamethylene imiuedithiocarbamate inhibits oxidation" of ethylbenzene (temperature 75°; STOICHIOM~TRIC

COEFFICIENTS

O F IXffHIBITIOlq ( f ) AT A TEMPERATURE

:FOR N I C K E L

C A R B A I ~ & T E S I1~ CUM:EN3~

OF 60 °

In air

Nickel carbamate

DiethyldithioDibutyldithio-

In oxygen according according to the to the according according according according trans transto v to T to We toW e formed formed curve curve 0. 6 1.2

1.3 0.5

1.0 1.4

0. 7 0.9

0. 8 0. 9

1.0 1.0

* 2~otv. T h e a v e r a g e a c c u r a c y o f a v e r a g e results is ± 2 0 % .

[ A I B N ] = 4 . 9 × 1 0 -3 moie/1., interacting with free radicals (f=0.6)). Values similar to ours were obtained [12] for f a n d kc, n a m e l y f is 1 for zinc diethyldithiocarbamate with initiated oxidation of tetralin at 30 ° a n d k e is approxim a t e l y 1.3 × 10a 1/mole.see. The mechanism of interaction of carbamates with RO~, apparently, involves electron transfer from the sulphur a t o m of carbam a t e to R0~ to form a peroxy anion [2, 7, 10]; this transfer m a y be considerably alleviated b y the presence of a metal a t o m and a covalent bond between this a t o m a n d the sulphur a t o m [7, 11]

RO '+

S CN R',

nO- + n',NC/s'+

\S--Ni_S / R,,NC j S

SNCN ' -

\S_Ni_S /

+ R' Nc~ S .

R,

--$ s"

S--

, + Ni+2.

SUMMARY

\ 1. Nickel diethyl- and dibutyldithiocarbamates react with cumyl hydroperoxide in chlorobenzene by a second order reaction with the same rate constant k = 5 X l0 s exp (--5000/RT) 1./mole. sec. One carbamate molecule breaks down six hydroperoxide molecules. The reaction takes place homolytically, the probability of volumetric radical yield varying between 0.2 at 20 ° a n d 0.7 at 70 ° . 2. Carbamates break the chains b y the reaction with cumyl peroxy radicals; a t 60 ° kk ( R O ~ - c a r b a m a t e ) - ~ 2 . 3 × 1 0 3 1./mole.see. The stoichiometric coefficient of inhibition is 1.

Inhibition mech~rtism of nickel dialkyldithioearb~mates

33

REFERENCES 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14.

L. A. BROOKS, Rubb. Chem. i Teehnol. 36, 887, 1963 T. COLILOUGH and J. I. CUNNEF2ff, J. Chem. See. 12, 4790, 1964 J. D. HOLDSWORTH, G. SCOTT and D. WILLIAMS, J. Chem. See. 12, 4692, 1964 ¢. COPPING and N. URI, Disc. F a r a d a y See. 46, 202, 1968 J. C. N. CHTENI and C. R. BOSS, J. Polymer Sei. A - l , 10, 1579, 1972 I. V. S ~ A N T S , IV[. A. DZYUBINA, N. V. VOYEVODA, V. V. SHER and P. I. SANIN, ~eftekhimiya 13, 749, 1973 S. K. IVANOV and V. S. YURITSYN, ~Toftekhimiya 11, 99, 1971 I. V. SH~H1YANTS, Ye. I. MA~KOVA, N. V. VOYEVODA, V. V. SKER and P. L SANIN, ~Teftekhimiya 11, 910, 1971; 9, 616, 1969 I. V. SH~HIYANTS, l~I. A. DZYUBINA, V. V. SH~R and P. I. SANI~, ~Teftokhimiya 13, 570, 1973 V. G. VINOGRADOVA and Z. K. MAIZUS, Kinotika i kataliz 13, 298, 1972 A. N. ZVEREV, V. G. VI~OGRADOVA and Z. K. MAIZUS, Izv. AN SSSR, Sot. khim., 2437, 1973 J. A. I~OWARD, Y. AHKATSU, J. Id[. B. ¢HENIER and K. U. INGOLD, Canad J. Chem. 51, 1543, 1973 L S. GAPONOVA, G. B. FEDOTOVA, V. F. TSEPALOV, V. F. SI~UVALOV a n d Ya. S. LEBEDEV, Kinetika i kataliz 12, 1137, 1971 Ye. T. DENISOV, Konstanty skorosti gomolitieheskikh zhidkofaznykh roaktsii (Rate Constants of Hemolytic Liquid-phase Reactions). Nauka, Moscow, 1971