Polymer International 42 (1997) 289 300

Liquid Natural Rubber as a Viscosity Modifier in Nitrile Rubber Processing

N. Radhakrishnan Nair,a * Sabu Thomasb & N. M. Mathewa
a Rubber Research Institute of India, Kottayam 686 009, Kerala, India b Mahatma Gandhi University, Priyadarsini Hills PO, Kottayam 686 650, Kerala, India (Received 12 March 1996 ; revised version received 24 July 1996 ; accepted 23 August 1996) Abstract : The use of thermally depolymerised liquid natural rubber as a plasticiser in nitrile rubber has been investigated with special reference to the rheological aspects. A comparative evaluation has been made with dibutylphthalate, the ester-type plasticiser commonly used with nitrile rubber. The study revealed that the use of liquid natural rubber as a plasticiser and viscosity modier improved the processability of the compounds and also enhanced many of the vulcanisate properties. Key words : liquid natural rubber, plasticiser, nitrile rubber, viscosity modier.

INTRODUCTION Oligomers are used as process aids in various compounded rubbers and plastic formulations. The advantages over conventional oil and ester-type plasticisers are their easier addition, improvement in properties for oil-resistant applications, where the plasticiser is not to be leached by the surrounding oil medium, enhancement of physical properties due to their polymeric nature, and better processability. The heterogeneous nature of rubber compounds, the strong interaction between the various formulation components, the viscoelastic nature and poor understanding of ow behaviour of such complex materials, make the processing of rubber compounds a complex procedure. Changes in the recipe are often made, either to meet new specications for nished products or to compensate for variations in properties or quality of raw materials. Improvement in rubber processing properties has become necessary owing to the need for high performance rubber formulations and because of competitiveness with the introduction of new challenging materials such as thermoplastic elastomers. Modication in bulk viscosity of compounds can result in easier processing, fewer rejects and reduction in energy consumption. An improvement of ow behaviour can allow higher output, or simpler operation, through greater latitude in mould design.
* To whom all correspondence should be addressed.

Annual consumption of natural rubber (NR) is about ve million tons and it remains as a major elastomer with unique properties for certain applications, despite continuous progress with synthetic polymers. The production potential of NR is larger than the present consumption and, because of its natural photosynthetic origin, it is an interesting renewable source of raw polymeric materials. The origin of natural rubber is unique
TABLE 1. Properties of LNR Ash content (%) Volatile matter (%) Specific gravity Colour Brookfield viscosity (cps at 38C) 1 M w 1 M n Polydispersity index 006 01 092 light brown 112 000 44 210 5 929 748

TABLE 2. Properties of NBR Volatile matter (%) Ash (%) Soap (%) Antioxidant (%) Bound ACN content (%) Mooney viscosity ML(1 4)=100C 075 10 005 10 33 50

289
Polymer International 0959-8103/97/$09.00 ( 1996 SCI. Printed in Great Britain

290

N. Radhakrishnan Nair, S. T homas, N. M. Mathew

TABLE 3. Formulation of mixes Compound NBR ZnO Stearic acid Liquid natural rubber Dibutyl phthalate Accelerator CBS Sulphur C 100 5 3 15 15 L5 100 5 3 5 15 15 L10 100 5 3 10 15 15 L15 100 5 3 15 15 15 L20 100 5 3 20 15 15 L25 100 5 3 25 15 15 D5 100 5 3 5 15 15 D10 100 5 3 10 15 15 D15 100 5 3 15 15 15 D20 100 5 3 20 15 15 D25 100 5 3 25 15 15

among the widely used polymers, because it occurs as the polymer itself, while others are manufactured from their respective monomers, mostly of petroleum origin. In order to obtain oligomers of natural rubber, depolymerisation methods must be resorted to. Synthetic rubber oligomers are obtained by controlling the processing conditions during the polymerisation process. Depolymerisation of natural rubber at dierent stages has been extensively studied in the past. The rst commercial production of depolymerised natural rubber in the USA was by depolymerising solid rubber.1 A pilot plant for latex stage depolymerisation was set up on the Ivory Coast.2 Production of liquid natural rubber by the combined eect of mechanical, chemical and thermal energies has been reported by Claramma et al.3 Nitrile rubber is generally difficult to process owing to the high viscosity of the compounds. The use of low molecular weight polymers in rubber compounding has been reported by Royale4 and Hashimoto.5 Thermally depolymerised natural rubber as a reactive plasticiser in nitrile rubber has been studied by Radhakrishnan Nair et al.6 Compounding nitrile rubber with furfuryl alcohol as a plasticiser has been reported by Magrupov et al.7 The use of latex-stage depolymerised natural rubber, using a phenyl hydrazine redox system as a bulk viscosity modier in rubber compounding, has been reported by Le Blanc et al.8 The present work on the plasticisation of nitrile rubber using liquid natural rubber produced by thermal depolymerisation was undertaken with a view to increasing the technoeconomic potential of natural rubber and gaining an insight into easier processing of nitrile rubber.

sample in the medium viscosity range was used for plasticising nitrile rubber. The physical properties of LNR are given in Table 1. Nitrile rubber of medium acrylonitrile content (grade N-339, manufactured by Synthetics and Chemicals, Bereilly, India) was used for the preparation of the compounds. Specication parameters of the nitrile rubber sample are given in Table 2. Dibutyl phthalate (DBP) used for compounding was laboratory grade N-dibutyl phthalate supplied by Sysco Chem. Industries, Bombay. Other chemicals used were of industrial purity.

Preparation of compounds
A nitrile rubber formulation with N-cyclohexylbenzthiazylsulphenamide (CBS) sulphur cure system was used. An unlled recipe was selected in order to observe the changes without the inuence by ller. Mixing was done on a laboratory-scale two-roll mill with a friction ratio of 1 : 125. A compound containing no plasticiser was taken as standard. Plasticiser content in the range 0 to 25 phr loading was investigated. Recipes of the dierent formulations are given in Table 3.

Testing
For the determination of cure time, an oscillating disc rheometer was used. The viscosity of the compounds was determined using a Mooney viscometer (SPRI Ltd, Negretti Automation, UK). A capillary rheometer attached to a Zwick UTM model 1474 was used for the rheological studies. The dimensions of the capillary rheometer are given in Table 4. The sample for testing was placed in the barrel, which was maintained at the
TABLE 4. Physical dimensions of capillary rheometer Diameter of capillary (mm) Length of capillary (mm) Diameter of plunger (mm) Diameter of cylinder (mm) 127 5008 1599 1600

EXPERIMENTAL

Materials
Thermally depolymerised liquid natural rubber (LNR) was used for the studies. The starting material for the preparation was ISNR-5. Depolymerisation was carried out in a thermal depolymeriser. A liquid natural rubber

POLYMER INTERNATIONAL VOL. 42, NO. 3, 1997

L iquid natural rubber as a viscosity modier

TABLE 5. Properties of compounds Compound C L5 L10 L15 L20 L25 D5 D10 D15 D20 D25 Scorch time (min) 525 525 550 675 700 700 525 550 575 600 600 Cure time (min) 215 240 225 230 225 250 195 230 220 220 195 Mooney viscosity 38 32 28 20 15 10 33 25 21 18 15

291

test temperature. The sample was forced down the capillary using the plunger attached to the xed cross-head. After a warming-up period of 3 min, the melt was extruded through the capillary at preselected speeds of the cross-head. The melt height in the barrel before extrusion was kept the same in all experiments and the machine was operated to give ten dierent plunger speeds. Each plunger speed was maintained until the recorded force stabilised, before changing to the next speed. Forces corresponding to specic plunger speeds were recorded. The force and the cross-head speeds were converted into corresponding apparent shear stress (q ) and the shear rate (c ) at the wall by using w w the following equations involving the geometry of the capillary and plunger : q \ F/4A (l /d ) (1) w p c c c \ [(3n ] 1)/4n]32Q/n dc3 (2) w where F is the force applied at a particular shear rate, A is the cross-sectional area of the plunger, l is the p c length of the capillary, d is the diameter of the capilc lary, Q is the volumetric ow rate, n is the ow behaviour index dened by d(log c )/d(log c ), c is the w wa wa apparent shear rate at the wall, and c is the actual w shear rate at the wall. The value of n was determined by regression analysis of the values of q and c obtained w wa from the experimental data. The shear viscosity g was calculated as : g \ q /c (3) w w The melt viscosity of the compounds was also calculated at shear rate 333 s~1 and 120C. The principal normal stress dierence, q [ q , was calculated from 11 12 the formula : [ q \ 2 ] q [2(d /d )6 [ 2]1@2 (4) 11 12 w e c where q is the shear stress at the wall, d is the diamw e eter of the extrudate and d is the diameter of the capilc lary. Recoverable elastic shear stress, SR, was calculated q
POLYMER INTERNATIONAL VOL. 42, NO. 3, 1997

from the formula : [ q )/2q (5) 11 12 w Elastic shear modulus, G, was calculated using the equation : G \ q /SR (6) w The samples extruded through the capillary were collected at selected shear rates and their diameters were determined using an optical microscope. Die swell values were calculated from the formula : die swell (%) \ [(d [ d )/d ] ] 100 (7) e c e The free vibration method on a Yerzely oscillograph has been used for measuring the important engineering properties of the vulcanisates. Static modulus (I) has been determined from the slope of the loading curve. Resilience (Y R) has been expressed as a percentage, determined as the average computed from the second and third half cycles : Y R (%) \ [(CD/BC) ] (DE/CD)] ] 50 (8) SR \ (q

where BC is the vertical distance of the rst half cycle of the damped sinusoidal wave, CD is the vertical distance of the downstroke of the second cycle and DE is the vertical distance of the upstroke of the second cycle. Natural frequency was determined by counting a convenient number of complete cycles, measuring the horizontal distance JK traversed by this number of cycles, X, along the centre axis. The frequency was calculated for a drum of 15 inch circumference and 4 rpm rotation, using the equation : f (Hz) \ X/JK (9)

The loss factor was calculated according to the method suggested by Scott9 from the formula : A /A \ e~1@2n tan d (10) 3 2 where A and A are the amplitudes of the third and 3 2 second vibrations and tan d is the loss factor.

292

N. Radhakrishnan Nair, S. T homas, N. M. Mathew superauto stigma and superauto focus. The fractured face of the extrudate sample was coated with gold under vacuum and photographed at appropriate magnications.

RESULTS AND DISCUSSION

Cure characteristics
Table 5 shows the processing parameters of the nitrile rubber compounds. Scorch time of the compounds indicated that the increased addition of dibutyl phthalate to nitrile rubber compounds had little eect. Slightly higher cure times were recorded on the rheometer for the LNR- containing compounds. Figure 1 shows the rheographs of compounds plasticised with 10 and 20 phr of DBP and LNR in comparison with the unplasticised stock. The torque showed a steady decrease with increased loading of the plasticisers. Comparing DBP and LNR, the latter showed more reduction of torque, especially at higher loadings. Mooney viscosities of the compounds are also given in Table 5. From the data it can be inferred that addi-

Fig. 1. Eect of plasticiser on cure.

Scanning electron microsopy studies were carried out on a general purpose scanning electron microscope (Hitachi model S-200). The equipment was one of the intelligent series with features such as high resolution,

Fig. 2. Shear stress versus shear rate at 90C. POLYMER INTERNATIONAL VOL. 42, NO. 3, 1997

L iquid natural rubber as a viscosity modier

293

Fig. 3. Viscosity versus shear rate at 90C.

tion of the plasticisers decreased the Mooney viscosity : the greater the quantity of the plasticiser, the greater was the reduction. Comparing the plasticisers, the inuence was more pronounced for LNR at higher loading.

respect to the shear rate axis. This can be represented by the power law equation q P cn (11)

Capillary rheometric studies

Shear stress versus shear rate. Shear stress values at different shear rates for the compounds containing LNR and DBP plasticisers and the control compound without any plasticiser were measured at 90C. The results are given in Fig. 2. The shear stress values at all shear rates were highest for the control and lowest for the compounds containing LNR, indicating more eective plasticiser action by LNR then by DBP. Comparison of the shear stresses at dierent shear rates indicated a shear rate thinning trend with increasing shear rate. This indicates pseudoplastic behaviour. The random ground state structure is progressively broken down with increase in shear rate and a shear oriented structure is formed, which oers reduced resistance to the force eld. Pseudoplasticity represents the behaviour of a uid which, when subjected to a shear rate from zero upwards, shows a convex curvature with
POLYMER INTERNATIONAL VOL. 42, NO. 3, 1997

where n is the ow behaviour index. Another possibility is that the curvature of the shear stress/shear rate curve is due to some limiting mechanical degradation occurring in the capillary at higher shear rates. The eect of shear rate on viscosity of the compounds containing 15 phr of LNR, DBP and the control compound are shown in Fig. 3. The viscosities at low shear rates were highest for the unplasticised compound, followed by DBP and LNR. As the shear rate increased, a drop in viscosity was noticed in all cases. The pseudoplastic nature of the compounds can be attributed to the structured state of entanglement at zero shear rate being progressively destroyed upon increasing shear. The LNR-containing compound showed the minimum viscosity. Eect on ow index. Figure 4 shows the change in ow index with increasing plasticiser content. In general, the ow index was higher for LNR compounds at all loadings. In fact the values increase linearly with LNR

294

N. Radhakrishnan Nair, S. T homas, N. M. Mathew

Fig. 4. Inuence of plasticiser on ow index.

Fig. 5. Viscosity and ow index versus temperature. POLYMER INTERNATIONAL VOL. 42, NO. 3, 1997

L iquid natural rubber as a viscosity modier

295

Fig. 6. Log viscosity versus 1/T .

content. In the case of DBP the increase in the ow index was nominal, indicating little impact of the plasticiser on the pseudoplastic ow behaviour of the compounds. LNR-containing compounds revealed that plasticiser loading has a more profound eect on the ow behaviour. For materials with pseudoplastic behaviour, the value of n in eqn (11) should be below unity. When high shear rates are operating on a polymer melt, the chain segments of the polymer molecules are aligned along the stream lines and thus increase the ow. The presence of LNR also inuences these orientations and contributes towards the ow behaviour. Inuence of temperature. Figure 5 shows the ow index of the compounds containing 15 parts of LNR and DBP as a function of temperature. With increase in temperature, the ow index increased in both cases. A consistent reduction in viscosity values was also observed with temperature. At all temperatures, the DBP compounds showed higher viscosities. A small levelling trend was observed at elevated temperatures, which was more prominent for LNR compounds and was attributable to scorching. Increase in temperature increases the Brownian motion in polymeric liquids and soft solid materials and the free volume also increases. Increase in molecular mobility enhances the ow behavPOLYMER INTERNATIONAL VOL. 42, NO. 3, 1997

iour. Viscosity temperature relations are governed by the Arrhenius equation : g \ A e~E@RT (12)

where E is the activation energy of ow and T is the absolute temperature. In order to obtain the activation energy, log viscosity was plotted against 1/T at a shear rate of 333 s~1 (Fig. 6). The slopes of the lines were calculated and the activation energy values of LNR and DBP samples were found to be 1128 and 846 kcal/ mol~1 respectively. The dependence of ow properties on temperature is of interest. Temperature is an eective means of controlling melt viscosity in processing operations. The optimisation of factors such as time, cost and money is necessary, and a control on thermal energy is required to save the polymer from degradation due to excessive heat and to provide long life to the products. Die swell. The die swell values of the compounds at different shear rates, with varying loading of the plasticisers from 0 to 20 phr, are given in Fig. 7. The temperature at which the extrudates were collected was 120C. It is observed that the die swell values decrease with increasing shear rate. A similar trend is observed with plasticiser loading. Comparing the plasticisers, the LNR-containing compounds exhibited die swell to a

296

N. Radhakrishnan Nair, S. T homas, N. M. Mathew

Fig. 7. Die swell at dierent shear rates at 120C.

lower extent than the corresponding DBP compounds. At higher shear rates the dimensional stability of LNRcontaining compounds was remarkably better. The surface nish of the extrudates was also good (Fig. 8).

Eect of shear rate on melt elasticity. Melt elasticity of the compounds was calculated from the extrusion parameters at a shear rate of 333 s~1 and a temperature of 120C. The results are depicted Fig. 9. The principal normal stress dierence, q [ q , decreased with 11 12 increased loading of the plasticiser. The primary or principal normal stress dierence of viscoelastic liquids represents a combination of elastic and viscous eects. A large dierence in the values of q [ q with 11 12 increase in plasticiser content indicated the inuence of plasticiser loading on the viscosity of the compounds. Recoverable elastic stress showed a decreasing trend with increasing plasticiser content. Minor inections in the values of shear modulus, G, were also observed. Morphology of the extrudates. Morphology studies were conducted on fractured faces of the extrudate samples rendered brittle cryogenically. A comparison of the scanning electron micrographs (Fig. 10) revealed good compatibility of LNR in the nitrile rubber matrix. No evidence of phase separation could be observed and the materials were found to be homogeneous at the 20 km
POLYMER INTERNATIONAL VOL. 42, NO. 3, 1997

Fig. 8. Photograph of extrudates.

L iquid natural rubber as a viscosity modier

297

Fig. 9. Melt elasticity at 333 s~1 and 120C. =, q [ q , ], recoverable elastic shear stress ; \, shear modulus, G. 11 12

scale. Good distribution was attained in the case of both LNR and DBP. No dierence was observed between the phase interactions of the two plasticiser systems with nitrile rubber. This suggests that LNR is an eective plasticiser for NBR compounds. V ulcanisate properties. Stressstrain properties of the vulcanisates containing plasticiser loaded to dierent levels were measured and are given in Fig. 11. Modulus values of the compounds at 300% elongation showed a decreasing trend with increase in plasticiser content. Tensile strength of the compounds showed a marginal reduction on addition of both the plasticisers at 5 to 10 phr levels, but remained constant thereafter. Elongation at break showed a decreasing trend initially (5 to 10 phr) followed by an increase. LNR at high loadings showed higher elongation values, attributable to the polymeric nature of the plasticiser. In agreement with the general concept of plasticisation, both LNR and DBP made the rubber compounds softer with increasing plasticiser content. However, this decrease was less for LNR, indicating its use for making relatively harder vulcanisates. Dynamic mechanical properties determined by using a Yerzely oscillograph are given in Fig. 12. Yerzely resilience values were in general higher for LNRcontaining compounds than for DBP-containing compounds. This may be attributed to the polymeric
POLYMER INTERNATIONAL VOL. 42, NO. 3, 1997

contribution of LNR, which is not present for DBP. Correspondingly higher values of loss factor were recorded for the DBP compounds than for the LNR compounds. The fundamental frequencies of the samples were comparable, within minor experimental uctuations. The static modulus showed a slightly decreasing trend with increasing plasticiser content. Evaluation of superposition shift factor. The experimental ow curves for log c versus q at temperatures of 90 to w w 120C were constructed for the nitrile rubber compounds. The method of superposition has been applied by the arbitrary choice of a reference temperature. The values of the superposition shift factors were obtained by choosing shear rate values at 100 and 1000 s~1 on the reference temperature ow curve and shifting the corresponding points (constant shear rate) on the ow curves for other temperatures to coincide with this shear stress. The values of a were calculated from the T equation : a \ q /q (constant) c (13) T (ref) (T) w where a is the shift factor, q is the shear stress at T (ref) reference temperature and q (constant) is the shear (T) stress at the particular study temperature. The average values of shift factors for the compounds were calculated10 and plotted as a function of temperature in Fig. 13. Using the appropriate average

298

N. Radhakrishnan Nair, S. T homas, N. M. Mathew

Fig. 11. Physical properties of vulcanisates. =, Modulus at 300% elongation (MPa) ; ], tensile strength (MPa) ; \, elongation at break ]100% ; K, hardness ]10 (Shore A).

SUMMARY AND CONCLUSIONS The eect of liquid natural rubber (LNR) in a nitrile rubber gum recipe was studied. It was compared with dibutyl phthalate (DBP), which is a popular ester-type plasticiser used with nitrile rubber. Cure characteristics of the compounds were comparable. LNR at higher loading showed a lower torque value on a rheograph than DBP. Studies on shear stressshear rate relations

Fig. 10. Scanning electron micrographs of extrudates : (a) control ; (b) L15 ; (c) D15.

values of the shift factors, the master ow curve at 90C was constructed and is given in Fig. 14. The shear rate temperature superposition method can be used for predicting the viscosities of the polymer melts at any required temperature by determining the viscosity at a reference temperature.

Fig. 12. Dynamic properties of vulcanisates. =, Resilience (%), ], tan d ; \, I (kg m2) ; K, Frequency ]10 (H2) ; x, hysteresis (%). POLYMER INTERNATIONAL VOL. 42, NO. 3, 1997

L iquid natural rubber as a viscosity modier

299

Fig. 13. Shift factor a versus temperature. T

Fig. 14. Superposition master curve. POLYMER INTERNATIONAL VOL. 42, NO. 3, 1997

300 using a capillary rheometer showed that both DBP and LNR decreased the compound viscosities with increased loading. Capillary rheometer measurements indicated the pseudoplastic nature of the compounds. The LNR compounds showed lower stress values than DBP compounds at equal shear rates. LNR acts as an eective plasticiser for nitrile rubber and aids the moulding process. The ow behaviour index for DBP showed steadier values than that for LNR on increase in their weight percentage. Temperature had a profound eect on the viscosity of the compounds. The activation energy of ow of LNR compounds was higher than that of DBP compounds. Compounds containing LNR resulted in extrudates with less swell and good surface nish. Morphological studies revealed compatibility of LNR with nitrile rubber. LNR compounds in general showed improvement in elongation and lower hardness reduction due to plasticiser. Dynamic mechanical properties were comparable. A master ow curve for the

N. Radhakrishnan Nair, S. T homas, N. M. Mathew compounds at 90C was also constructed. In general, LNR was found to be an eective plasticiser and a good processing aid in nitrile rubber compounding.

REFERENCES
1 Hardman, K. V. & Lang, A. J., Rubber Age, 66 (1950) 419. 2 Allet D. A. & Lemoine, A., Proceedings of the International Conference on L iquid Natural Rubber, Abidjan, 1986, p. 33. 3 Claramma, N. M., Radhakrishnan Nair, N. & Mathew, N. M. Indian J. Natural Rubber Res. 4 (1991) 1. 4 Royale, E., European Patent Appl. 401193 A1, 1990. 5 Hashimoto, G., Japanese patent No. 91100048 A2, 1991. 6 Radhakrishnan Nair, N., Claramma, N. M., Baby Kuriakoe & Mathew, N. M. Indian J. Natural Rubber Res. 2 (1991) 81. 7 Magrupov, F. A., Abdurashidov, T. & Gorbunova, M. I. USSR Patent SU 1558934, 1990. 8 Le Blanc, J. L., De Livonniere, H. P. & Boccasio, G. Prog. Rubber T echnol., 8 (1992) 79. 9 Scott, J. R. in Physical T esting of Rubbers. Maclaren, New York, 1965, p. 180. 10 Anand, J. S., Int. Plast. Eng. T echnol., 1 (1994) 25.

POLYMER INTERNATIONAL VOL. 42, NO. 3, 1997