zeitschrift kunststofftechnik journal of plastics technology · verheyen, giesen, heim curing...

4Autor Titel (gegebenenfalls gekürzt)

© Carl Hanser Verlag Zeitschrift Kunststofftechnik / Journal of Plastics Technology 13 (2017) 4

13 (2017) 4 eingereicht/handed in: 27.07.2017 angenommen/accepted: 14.09.2017

Fabian Verheyen M.Sc.; Dr.-Ing. Ralf-Urs Giesen; Prof. Dr.-Ing. Hans-Peter Heim Institut für Werkstofftechnik – Kunststofftechnik, Universität Kassel

Influence of different types and amounts of crosslinking agent on the curing process of silicone rubber In this investigation, a high-consistency rubber was mixed with five different types and amounts of curing agent. After mixing, the raw material was characterised using a rubber process analyser. The results led us to the conclusion that silicone rubber could be crosslinked with a wide range of curing agents based on organic peroxides. A large amount of curing agent is not necessary to achieve higher values in terms of the elastic torque, or to accomplish a shorter processing time. It was able to show that the crosslinking density according to δ S’ is on a similar level throughout a range of curing temperatures.

Einfluss verschiedener Vernetzungssysteme und Vernetzeranteile auf das Vulkanisations-verhalten von Silikonkautschuk In dieser Studie wurde ein Festsilikonkautschuk mit fünf verschiedenen Vernetzungssystemen, welche auf organischen Peroxiden basieren, und verschiedenen Mengenanteilen gemischt. Nach dem Mischen wurde das Vulkanisationsverhalten der Polymermischungen mit Hilfe eines Rubber Process Analyzer untersucht. Die Ergebnisse zeigen, dass Festsilikonkautschuk mit einer Vielzahl an Vernetzungssystemen vulkanisiert werden kann. Ein größerer Anteil an Vernetzer ist nicht notwendig, um den elastischen Anteil zu vergrößern oder die Vulkanisationszeit zu verkürzen. Es konnte zudem gezeigt werden, dass sich die Vernetzungsdichte über einen großen Temperaturbereich auf einem ähnlichen Niveau bewegt.

archivierte, peer-rezensierte Internetzeitschrift archival, peer-reviewed online Journal of the Scientific Alliance of Plastics Technology

Zeitschrift Kunststofftechnik Journal of Plastics Technology www.kunststofftech.com · www.plasticseng.com

© 2

017

Car

l Han

ser

Ver

lag,

Mün

chen

w

ww

.kun

stst

offte

ch.c

om

Nic

ht z

ur V

erw

endu

ng in

Intr

anet

- un

d In

tern

et-A

ngeb

oten

sow

ie e

lekt

roni

sche

n V

erte

ilern

.

Verheyen, Giesen, Heim Curing Process Of Silicone Rubber

Journal of Plastics Technology Zeitschrift Kunststofftechnik 13 (2017) 4 234

Influence of different types and amounts of crosslinking agent on the curing process of silicone rubber

F. Verheyen, R.U. Giesen, H.P. Heim

1 INTRODUCTION

Silicone rubber has a unique molecular structure, which consists of alternating silicone and oxygen bonds. This special structure creates organic and inorganic properties, and is responsible for the excellent heat resistance, chemical resistance, and flexibility of this material when exposed to a wide range of temperatures [1, 2, 3, 4]. Silicone rubber consists of different types of polysiloxanes, e.g. polydimethylsiloxane (=PDMS). The process of curing transfers the raw polymer to an elastomeric, three-dimensional network [1, 4]. Owing to these properties, silicone elastomers are well-suited for applications in the automotive field or in the medical industry [3, 4, 5]. The group of heat-cured silicone rubber is divided further into the group for liquid silicone rubber (=LSR) and the one for high-consistency silicone rubber (=HCR). In contrast to LSR, HCR has longer molecule chains, and is usually cured using organic peroxides [1, 6]. Organic peroxides have one or more oxygen-oxygen bonds [7]. The peroxide degrades into free radicals, which eliminate hydrogen atoms from the side groups of the PDMS, and form chemical bonds on the backbone of the silicone rubber [6, 8]. In addition to the type of silicone rubber that was successfully crosslinked, the main differences between peroxides include the types of free radicals, the energy levels of these radicals, and the half-life of the peroxide itself. The time at which one half of the peroxide decomposes at a given temperature is called half-life. The curing reaction is notably faster when the half-life is shorter. [7, 8] Examples of the mechanism of the radical reaction during curing are shown in [2, 4]. The vulcanisation causes the development of cleavage products. These cleavage products should be removed before usage by means a post-curing process, which is carried out at 200 °C for 4 h [4, 5, 9]. It is well known that the amount of curing agent has a fundamental influence on the processing conditions in addition to the type of crosslinking agent, and also on the properties of the silicone rubber [6]. The processing parameters are influenced by the half-life of the peroxide used. Decomposition is temperature-sensitive, and, accordingly, the crosslinking reaction occurs at a different rate. The degree of crosslinking is proportional to the concentration of the crosslinking agent [7, 8]. As the amount of crosslinking agent increases, the crosslinking density increases as well [8, 10, 11]. The degree of crosslinking

© 2

017

Car

l Han

ser

Ver

lag,

Mün

chen

w

ww

.kun

stst

offte

ch.c

om

Nic

ht z

ur V

erw

endu

ng in

Intr

anet

- un

d In

tern

et-A

ngeb

oten

sow

ie e

lekt

roni

sche

n V

erte

ilern

.

http://www.dict.cc/englisch-deutsch/development.html



affects the hardness, ultimate elongation, and other properties, like the tear strength [7, 8, 11]. A rubber process analyser is well suited for the characterisation of the curing behaviour of elastomers. There are several investigations in literature which investigate the curing characteristics of rubber when using a rubber process analyser (=RPA), e.g. [12-16]. Zhang et al. [16] found out that the crosslinking density of natural rubber decreases as the curing temperature increases. The optimal curing time decreases with rising curing temperature. In consequence of a lower cross-link density the modulus are on a lower level. Dick found out that the curing time reduces in correlation to the rising amount of peroxide. [7] There is less data available about the influence of crosslinking agents based on organic peroxides and their influence on the curing behaviour of silicone elastomers. Only a range of an amount is provided in the material data sheet of the producer, and there is no statement concerning the influence of additional crosslinking agent. In this investigation, a high-consistency rubber was mixed with five different, typically used crosslinking agents that are based on organic peroxides. Furthermore, the amount of crosslinking agent was varied in order to quantify the effect on the curing behaviour. After mixing, all batches were investigated using a rubber process analyser. The goal of this investigation is to analyse the curing process of silicone rubber by using different kinds of organic peroxides. Apart from the type of silicone rubber employed, the influence of the curing temperature and the amount of crosslinking agent on the curing process was investigated. Profound knowledge of the curing process is needed to prevent premature curing e.g., in the mould. By using a customized processing temperature, the producer can reduce the amount of energy required, which, respectively, reduces the cycle time.

2 EXPERIMENTAL

2.1 Materials In this study, the base polymer is a high consistency rubber called Silplus 60EX from the company Momentive Performance Materials Inc. (Leverkusen/ Germany), which is typically processed using extrusion [17]. Five different types of crosslinking agents were used, all of which were manufactured by Akzo Nobel GmbH (Amsterdam/Netherlands). They are usually used for crosslinking of silicone rubber. All crosslinking agents were mixed in silicone oil and delivered as a paste. These masterbatches reduce the mixing time and increase the dispersion of the crosslinking agent in the polymer. [7]

© 2

017

Car

l Han

ser

Ver

lag,

Mün

chen

w

ww

.kun

stst

offte

ch.c

om

Nic

ht z

ur V

erw

endu

ng in

Intr

anet

- un

d In

tern

et-A

ngeb

oten

sow

ie e

lekt

roni

sche

n V

erte

ilern

.



DClBP, DMB and DBP belong to the class of diacyl peroxides, and display shorter half-lives (see table 1) and free radicals with high energy levels. DCP and DMTB belong to the group of dialkyl peroxides. DMTB generates a mix of free radicals with high and low energy levels. The chemical structure of all of the employed peroxides is shown in figure 1. [7]

Figure 1: Chemical structure of the employed peroxides [18] DClBP and DMB were able to be cured without employing external pressure, e.g. by using hot air or infrared-radiation. The other curing agents were able to be cured under pressure in a hot press. Table 1 show some chosen properties of the used peroxides. No. Crosslinker

[-] Chemical Formula

[-] Abbr.

[-] Recommended

Curing Temperature

[°C]

Recommended Amount

[phr]

Half Life 1h

[°C]

Half Life 0.1h [°C]

Class of peroxide

[-]

1 Perkadox PD-50S-PS

Di(2,4-dichlorobenzoyl)

peroxide

DClBP 90 1.1-2.3 65 80 Diacyl- peroxide

2 Perkadox PM-50S-PS

Di(4-methylbenzoyl) peroxide

DMB 105 1.1-2.3 77 98 Diacyl- peroxide

3 Perkadox L-50S-PS

Dibenzoylperoxide DBP 105 0.7-1.4 91 132 Diacyl- peroxide

4 Perkadox BC-40S-PS

Dicumylperoxide DCP 175 --- 138 162 Dialkyl- peroxides

5 Trigonox 101-45S-PS

2,5-Dimethyl-2,5-di(tert-butylperoxy)

hexane

DMTB 175 --- 147 172 Dialkyl-peroxides

Table 1: Selected properties of the used crosslinking agents [7, 18]

© 2

017

Car

l Han

ser

Ver

lag,

Mün

chen

w

ww

.kun

stst

offte

ch.c

om

Nic

ht z

ur V

erw

endu

ng in

Intr

anet

- un

d In

tern

et-A

ngeb

oten

sow

ie e

lekt

roni

sche

n V

erte

ilern

.



2.2 Preparation All materials were mixed using the conical twin mixer CTM 25 from Colmec S.p.A. (Busto Arsizio/Italy) using the same mixing parameters.

2.3 Rubber Process Analyser After mixing, the different, uncured compounds were tested with a rubber process analyser D-RPA 3000 that was made by MonTech Werkstoffprüfmaschinen GmbH (Buchen/Germany). An RPA is well-suited for the investigation of the curing behaviour of rubber. The temperature-controlled testing chamber consists of two biconical cavities, which both have the same angle. A shear strain is applied to the specimen by oscillating the lower die sinusoidally, and the torque response of the material is measured. Owing to this, the rheological characterisations of gum and rubber compounds and the curing behaviour of these compounds were able to be investigated using an RPA. [6, 12-14, 19, 20]

Figure 2: The design of an RPA according to [21] For each measurement, approximately 6 g of the compound was placed between two testing films of polyethylene terephthalate (=PET) with a thickness of 23 µm, and was then put in the RPA. All formulae were tested at different temperatures. Isotherms were employed in a range from 180 °C to 90 °C according to [21]. The frequency and strain remained constant during testing (1.67 Hz and 7 %), and the testing time was 20 min. The elastic torque, the time to scorch, TC 90 and δ S’ were analysed. The elastic torque (S’) in [dNm] characterises the curing behaviour of the different compounds. The scorch time (TS2) in [s] is the time measured at the rotor start which is needed for an increase of 2 dNm above the torque of S’ min. TC 90 in [s] is the time needed to reach a curing level of 90 %. The time is calculated between S’ min and S’ max (see figure 3) [9]. Delta torque (δ S’) in [dNm] is calculated by the difference between S’ max and S’ min. It provides

© 2

017

Car

l Han

ser

Ver

lag,

Mün

chen

w

ww

.kun

stst

offte

ch.c

om

Nic

ht z

ur V

erw

endu

ng in

Intr

anet

- un

d In

tern

et-A

ngeb

oten

sow

ie e

lekt

roni

sche

n V

erte

ilern

.



information about the crosslinking density of the cured rubber. The larger δ S’, the higher the crosslink density of the system compared to [15, 16, 23]. Figure 3 summarizes the most important values.

Figure 3: Curing parameters according to [21, 22]

3 RESULTS AND DISCUSSION

3.1 Calculation of the Half-Life of Organic Peroxide The half-life can be calculated using the Arrhenius equation. According to [24, 25], the half-life was calculated using formulas 1 and 2 in correlation with different curing temperatures. In this equation, 𝑘𝑘𝑑𝑑 is the rate invariable for the initiator dissociation, A is the Arrhenius frequency factor, 𝐸𝐸𝑎𝑎 the activation energy for the initiator dissociation, R the general gas invariable, and T the absolute temperature in Kelvin.

𝑘𝑘𝑑𝑑[1𝑠𝑠] = 𝐴𝐴 ∗ 𝑒𝑒− 𝐸𝐸𝑎𝑎𝑅𝑅∗𝑇𝑇 (1)

𝑡𝑡0,5[𝑠𝑠] = ln(2)𝑘𝑘𝑑𝑑

(2)

© 2

017

Car

l Han

ser

Ver

lag,

Mün

chen

w

ww

.kun

stst

offte

ch.c

om

Nic

ht z

ur V

erw

endu

ng in

Intr

anet

- un

d In

tern

et-A

ngeb

oten

sow

ie e

lekt

roni

sche

n V

erte

ilern

.



The concentration of peroxide (𝑙𝑙1) at a given time is calculated using formula 3. Here, 𝑙𝑙0stands for the original concentration of peroxide, and t represents the time measured at the start of decomposition.

𝑙𝑙1 = 𝑙𝑙0 ∗ 𝑒𝑒−𝑘𝑘𝑑𝑑∗𝑡𝑡 (3)

Table 2 show the calculated half-lives of the used peroxides as a function of the different curing temperatures. All values over 1200 s (light grey) could not be detected by the following RPA measurements, because of the defined testing time of 1200 s. Diacyl peroxides have a much shorter half-life in contrast to dialkyl peroxides. As a result, the curing speed might be much faster at the same curing temperature.

Table 3 shows the rest of the concentration of peroxide at a defined time. The values were calculated according to formula 3 [25]. In the case of the dialkyl peroxides, undecomposed peroxide was still found in the polymer after 1200 s at a curing temperature of 150°C (DCP: 0.13 phr and DMTB: 0.28 phr). The diacyl peroxides were completely decomposed after this time. This indicates that the reaction process was completed in the case of the diacyl peroxides. Hence, the elastic torque S’ should be constant.

Temperature DClBP DMB DBP DCP DMTB [°C] [s] [s] [s] [s] [s] 180 0 1 1 54 106

170 0 2 2 124 256

160 1 4 4 294 645

150 2 8 9 727 1,697

140 4 17 19 1,876 4,677

130 8 39 44 5,074 13,560

120 19 94 103 14,438 41,499

110 45 236 252 43,387 134,640

100 117 624 650 138,298 465,269

90 318 1,741 1,765 469,891 1,721,439

Table 2: Calculated Half-Lives of the Used Peroxides in [s]

© 2

017

Car

l Han

ser

Ver

lag,

Mün

chen

w

ww

.kun

stst

offte

ch.c

om

Nic

ht z

ur V

erw

endu

ng in

Intr

anet

- un

d In

tern

et-A

ngeb

oten

sow

ie e

lekt

roni

sche

n V

erte

ilern

.



Time DClBP DMB DBP DCP DMTB [s] [phr] [phr] [phr] [phr] [phr]

TC90 0.00 0.00 0.05 0.20 0.22

600 0.00 0.00 0.00 0.23 0.35

1200 0.00 0.00 0.00 0.13 0.28

Table 3: Remaining concentration of peroxide in [phr] at a defined time, a curing temperature of 150°C, and an original concentration of 1 phr masterbatch

Figure 4 shows the decomposition of the peroxide as a function of the number of half-lives [24, 25]. Theoretically, total decomposition was unable to be achieved. Dick suggested in [7] that the minimum cure time be ten times a half-life. TC90 was reached after four times a half-life. Maier and Schiller [24] suggested six times to eight times a half-life for total decomposition of the peroxide. As a result, the elastic torque S’ should be constant.

Figure 4: Percentage of original decomposed peroxide as a function of the number of half-lives [7]

© 2

017

Car

l Han

ser

Ver

lag,

Mün

chen

w

ww

.kun

stst

offte

ch.c

om

Nic

ht z

ur V

erw

endu

ng in

Intr

anet

- un

d In

tern

et-A

ngeb

oten

sow

ie e

lekt

roni

sche

n V

erte

ilern

.



3.2 Static RPA-Measurements

The following results show the elastic torque of HCR as a function of the testing time. The curing temperature and amount of curing agent were modified. Depending on the used curing agent, the graph shows different appearances. Figure 5 and figure 6 show the elastic torque (S’) of the different materials at the same testing temperature of 150 °C and 170 °C respectively. DBP achieved the highest value in terms of the elastic torque (S’). DClBP, DMB and DBP (all Diacyl peroxides) show nearly the same course of their graphs. In contrast to that, DCP and DMTB (Dialkyl peroxides) achieve lower values in terms of the elastic torque. This was also able to be shown in the case of the longer time to scorch (TS 2) or TC 90, see table 4. The time to scorch depends on the temperature and type of curing agent. DClBP, DMB and DBP, which have a low half-life (see table 1), show significantly lower scorch times (7, 8 and 9 s) in contrast to DCP and DMTB, which have scorch time of 32 s and 43 s. In the case of diakyl peroxides, the entire amount of peroxide is not completely decomposed at 150°C after a testing time of 20 minutes. As a result, the elastic torque S’ still continues to increase (see figure 5). Diacyl peroxides have a shorter half-life. After approx. 200 seconds, S’ is on the same level. This indicates that the peroxide decomposed completely during the test (compare to the results of the calculated half-life). As the temperature rose (up to 170°C – see figure 6), the half-life reduced further and the curing speed continued to rise. As expected, the scorch time also reduced as the temperature rose. The level of S’ depends on the peroxide used. DBP generates free radicals with the highest energy level, and, accordingly, creates strong chemical bonding of the silicone elastomer (compare to [7]). All diacyl peroxides are non-vinyl-specific. There are reaction points located at the vinyl groups and the methyl groups of the polysiloxane. The potential tie points increase in correlation with the amount of crosslinking agent, resulting in a higher elastic torque. [6] For example, DMTB displays the lowest value in terms of S’. There are multiple reasons as to why this is the case. Due to it having the longest half-life, less peroxide is decomposed during the process than in the case of diacyl peroxides. Additionally, only 45% of the original peroxide in the masterbatch is present. Furthermore, DMTB generates a mix of high- and low-energy free radicals [7]. Owing to this, strong bonding does not occur, resulting in a lower elastic torque S’. Dialkyl peroxides are all vinyl-specific peroxides, meaning the chemical bonding only takes place at the vinyl groups of the polydimethylsiloxane. [6] As the temperature rises (up to 170°C), the S’ value of every compound reaches a plateau. The entire amount of peroxide is decomposed during the testing time of 20 min. The curing reaction is accomplished, and results in a stable elastic torque. According to formula 3, there is just 0.01 phr of DCP and 0.09 phr of DMTB after 600 s at 170°C. No DCP and 0.02 phr of DMTB remain in the polymer (after 1200 s).

© 2

017

Car

l Han

ser

Ver

lag,

Mün

chen

w

ww

.kun

stst

offte

ch.c

om

Nic

ht z

ur V

erw

endu

ng in

Intr

anet

- un

d In

tern

et-A

ngeb

oten

sow

ie e

lekt

roni

sche

n V

erte

ilern

.



Figure 5: Elastic torque as a function of time for the different curing agents with an amount of 1 phr @ 1.67 Hz, 7 % strain and at a curing temperature of 150 °C

Figure 6: Elastic torque as a function of time for the different curing agents with an amount of 1 phr @ 1.67 Hz, 7 % strain and at a curing temperature of 170°C

© 2

017

Car

l Han

ser

Ver

lag,

Mün

chen

w

ww

.kun

stst

offte

ch.c

om

Nic

ht z

ur V

erw

endu

ng in

Intr

anet

- un

d In

tern

et-A

ngeb

oten

sow

ie e

lekt

roni

sche

n V

erte

ilern

.



No. Cross-linker

Amount[phr]

Temp. [°C]

S’ min [dNm]

S’ max [dNm]

Scorch Time [s]

TC 90 [s]

1 DClBP 1.0 170 2.43 20.89 7 15

2 DMP 1.0 170 1.86 18.38 9 20

3 DBP 1.0 170 1.75 22.24 8 19

4 DCP 1.0 170 0.90 17.78 32 133

5 DMTB 1.0 170 0.86 16.24 43 178

Table 4: Selected results of the static measurements

Figure 7 shows the elastic torque (S’) at a curing temperature of 150 °C as a function of the amount of curing agent for DClBP as an example of the behaviour of diacyl peroxides. As the amount of curing agent increases, the elastic torque becomes higher (rising from 20 dNm to 28 dNm). Figure 8 depicts the results for the material behaviour of DCP (dialkyl peroxides). The course of each graph is nearly the same, but no consistent behaviour was detected. Depending on the amount of peroxide, a higher number of free radicals was built. The free radicals are responsible for a higher crosslinking density, which, in turn, results in a higher elastic torque S’.

Figure 7: Elastic torque (S’) as a function of time with different amounts of DClBP @ 1.6 Hz, 7 % strain and at a curing temperature of 150 °C

© 2

017

Car

l Han

ser

Ver

lag,

Mün

chen

w

ww

.kun

stst

offte

ch.c

om

Nic

ht z

ur V

erw

endu

ng in

Intr

anet

- un

d In

tern

et-A

ngeb

oten

sow

ie e

lekt

roni

sche

n V

erte

ilern

.



Figure 8: Elastic torque (S’) as a function of time with different amounts of

DCP @ 1.6 Hz, 7 % strain and at a curing temperature of 150 °C Table 5 and table 6 depict the decrease of the scorch time and TC 90 in correlation with the rising amount of curing agent. Depending on the curing agent and the recommended processing temperature, the values differ. The highest drop in time is between 1 and 1.5 phr.

Amount DClBP DMB DBP DCP DMTB

[phr] [s] [s] [s] [s] [s]

1.0 10 13 13 166 267

1.5 8 13 11 97 130

2.0 7 12 11 85 115

2.5 7 12 10 57 103

3.0 7 12 10 64 85

5.0 6 12 10 46 58

Table 5: Scorch time in [s] for the different curing agents at 150 °C

© 2

017

Car

l Han

ser

Ver

lag,

Mün

chen

w

ww

.kun

stst

offte

ch.c

om

Nic

ht z

ur V

erw

endu

ng in

Intr

anet

- un

d In

tern

et-A

ngeb

oten

sow

ie e

lekt

roni

sche

n V

erte

ilern

.



Amount DClBP DMB DBP DCP DMTB

[phr] [s] [s] [s] [s] [s]

1.0 21 39 31 587 744

1.5 19 29 29 469 510

2.0 18 30 31 412 468

2.5 18 29 30 493 429

3.0 18 30 31 421 395

5.0 18 30 31 400 299

Table 6: TC 90 in [s] for the different batches at 150 °C

Figure 9 shows the elastic torque (S’) of DClBP. The curing temperature was varied between 180 °C to 90 °C. As the curing temperature declined, the elastic torque decreased to a lower level. At 90 °C, there was no curing reaction of the material anymore (after 20 min of testing). DClBP, DMB and DBP cured until 90 °C was reached. DCP and DMTB cured up until 140 °C was reached. No curing reaction takes place below this temperature. As is visible in Table 2, the half-life of the used peroxide acts as a function of the temperature, and, in contrast, curing does not take place during the testing time if the temperature is too low. Table 7 shows the TC 90 as a function of the curing temperature. No curing took place for DMB under 100 °C. DCP and DMTB did not fully cure under 150 °C. The TC 90 time is strongly influenced by the curing temperature, like [12] and [16] have shown. As the curing temperature rose, the TC 90 became lower (compare to table 2). After a material-specific temperature was reached, no significant drop in the TC 90 occurred. In the case of DCP and DMTB, this temperature level is higher than the considered range of temperature. Accordingly, the curing temperature may be reduced in order to save energy costs. For example, there is no significant difference between the TC 90 (see table 7) or the elastic torque (see figure 9) above 130 °C.

© 2

017

Car

l Han

ser

Ver

lag,

Mün

chen

w

ww

.kun

stst

offte

ch.c

om

Nic

ht z

ur V

erw

endu

ng in

Intr

anet

- un

d In

tern

et-A

ngeb

oten

sow

ie e

lekt

roni

sche

n V

erte

ilern

.



Figure 9: Elastic torque (S’) as a function of time of 1 phr of DClBP with

different temperatures @ 1.6 Hz and 7 % strain

Temperature DClBP DMB DBP DCP DMTB

[°C] [s] [s] [s] [s] [s]

180 15 17 17 81 94

170 16 20 19 133 178

160 17 23 24 280 384

150 21 36 31 587 not full cured

140 25 50 46 not full cured not full cured

130 34 95 88 no cure no cure

120 56 213 194 no cure no cure

110 107 480 397 no cure no cure

100 232 906 742 no cure no cure

90 515 no cure 1136 no cure no cure

Table 7: TC 90 of 1 phr as a function of the curing temperature

© 2

017

Car

l Han

ser

Ver

lag,

Mün

chen

w

ww

.kun

stst

offte

ch.c

om

Nic

ht z

ur V

erw

endu

ng in

Intr

anet

- un

d In

tern

et-A

ngeb

oten

sow

ie e

lekt

roni

sche

n V

erte

ilern

.



Figure 10 and figure 11 show the results of the calculation of δ S’. The larger δ S’ is, the higher the crosslinking density of the system (see also [15, 16, 23]). Figure 7 shows δ S’ as a function of the amount of curing agent at the same curing temperature of 150 °C. Despite using the same base polymer, the level of δ S’ depends on the curing agent used. The highest values in terms of δ S’ were achieved by DBP. Only DClBP and DBP showed an increasing crosslinking density in correlation with the rising amount of curing agent. The remaining compounds were on the same level. In contrast to [10] and [11], we were able to show that the crosslinking density does not generally increase with the rising amount of curing agent. Instead, it depends on the type of curing agent used. As stated before, dialkyl peroxides are all vinyl-specific peroxides – that means the chemical bonding takes place only at the vinyl group of the Polydimethylsiloxane. A rising amount of curing agent has no effect on the crosslinking density if the amount of vinyl groups is equal to it [6]. In contrast to that, diacyl peroxides are non-vinyl-specific. More precisely, their chemical bonds are formed both at the vinyl groups and the methyl-groups of the polysiloxane. Owing to this, the potential bonding increase as the amount of crosslinking agent does [6]. For this reason, DMB should display an increasing crosslinking density as the amount of crosslinking agent increases, however, it does not.

Figure 10: δ S’ as a function of the amount of the curing agent @ 1.67 Hz, 7 % strain and at 150 °C

© 2

017

Car

l Han

ser

Ver

lag,

Mün

chen

w

ww

.kun

stst

offte

ch.c

om

Nic

ht z

ur V

erw

endu

ng in

Intr

anet

- un

d In

tern

et-A

ngeb

oten

sow

ie e

lekt

roni

sche

n V

erte

ilern

.



The results of δ S’ as a function of the curing temperature are shown in figure 11. All samples had an amount of curing agent of 1 phr. The optimal curing temperature in terms of the crosslinking density is 20 °C higher in contrast to the values provided in the data sheet of the material producer (see table 1). All graphs display a similar course – as the temperature rises, the δ S’ level increases. After reaching a material-specific temperature, δ S’ reaches a plateau. A further increase of the temperature leads to a decrease of δ S’ in the diacyl peroxides. Depending on the curing agent used, the optimal temperature and δ S’ can be on different levels. DBP achieve the highest values in terms of δ S’. During the curing reaction when diacyl peroxides are present, some acid cleavage products are created, which are responsible for the degradation of the polymer network. Moreover, the heat resistance of the silicone rubber in a closed system decreased when using diacyl peroxides like DClBP or DBP [6, 7]. The degradation of the polymer network might be the reason for the decreasing crosslinking density resulting in a lower delta S’ for DMB. DClBP and DBP, both of which are diacyl peroxides, show a slight decrease of delta S’ as the temperature increases.

Figure 11: δ S’ of different types of curing agent (1 phr) as a function of the curing temperature @ 1.67 Hz, 7 % strain

© 2

017

Car

l Han

ser

Ver

lag,

Mün

chen

w

ww

.kun

stst

offte

ch.c

om

Nic

ht z

ur V

erw

endu

ng in

Intr

anet

- un

d In

tern

et-A

ngeb

oten

sow

ie e

lekt

roni

sche

n V

erte

ilern

.



4 SUMMARY AND CONCLUSION

In this investigation, a high-consistency rubber was mixed with five different types and amounts of curing agent. After mixing, the raw material was characterised using a rubber process analyser. The calculated half-lives of the used organic peroxides illustrate the extremely temperature-sensitive behaviour of the different crosslinking agents. As a consequence of an increase of the curing temperature by 10°C, the half-life reduced at least 50% over time. Varying the processing temperature has a large influence on the curing behaviour of silicone rubber. The RPA measurements verify this significant temperature-sensitive behaviour. As the curing temperature rose, the scorch time and TC90 decreased in all batches. The curing speed depends on the peroxide used. Diacyl peroxides have a significantly shorter half-life in contrast to dialkyl peroxides, and, as a result, display a faster curing speed at the investigated curing temperatures (90-180°C). In correlation with the rising amount of crosslinking agent, the elastic torque increase. This effect also depends on the group of peroxides used. Diacyl peroxides are non-vinyl-specific, and, as a matter of fact, create more chemical bonds in the polymer network, which, in turn, lead to a higher crosslinking density. In contrast, dialkyl peroxides are vinyl-specific and have a defined crosslinking density that is dependent upon the original amount of vinyl groups in the polymer. A higher curing temperature affects the crosslinking density of the diacyl peroxide by causing degradation in the polymer network. We successfully proved that the half-life of the used peroxide, and, consequently, the curing temperature are significantly responsible for the curing behaviour of the silicone rubber. In the case of diacyl peroxide, a higher amount of crosslinking agent causes a higher crosslinking density. A profound knowledge of the curing process prevents early curing, and enables producers to tailor properties by employing different kinds of organic peroxides or different curing temperatures in order to reduce the amount of required energy or to reduce the cycle time, respectively.

5 ACKNOWLEDGMENTS The authors would like to thank Momentive Materials Inc. for supporting this investigation.

© 2

017

Car

l Han

ser

Ver

lag,

Mün

chen

w

ww

.kun

stst

offte

ch.c

om

Nic

ht z

ur V

erw

endu

ng in

Intr

anet

- un

d In

tern

et-A

ngeb

oten

sow

ie e

lekt

roni

sche

n V

erte

ilern

.



References

[1] Heiner, J.; Stenberg, B..; Persson, M.

Crosslinking of siloxane elastomers Polymer testing, 22 (2003) 3, S. 253-257. DOI: 10.1016/S0142-9418(02)00081-8

[2] Baquey, G.; Moine, L.; Degueil, M. et al.

Decomposition of Di-tert-butyl Peroxide in Siloxane: An Approach of the Free Radical Crosslinking of Silicones Macromolecules, 38 (2005) 23, S. 9571–9583. DOI: 10.1021/ma0511200

[3] Shit S.; Shah, P.

A review on Silicone rubber National Academy Science Letters, 36 (2013) 4, S. 355-365, DOI: 10.1007/s40009-013-0150-2

[4] Lopez, L.; Cosgrove, A.; Hernandez‐Ortiz, J.; Osswald, T.

Modeling the vulcanization reaction of silicone rubber Polymer Engineering & Science, 47 (2007) 5, S. 675-683, DOI: 10.1002/pen.20698

[5] N.N. Silicone – Chemie und Technologie Vulkan-Verlag, Essen, 1989

[6] Noll, W. Chemie und Technologie der Silicone Chemie-Verlag GmbH, Weinheim, 1968

[7] Dick, J. Rubber Technology Carl Hanser Verlag, München, 2009. DOI: 10.3139/9783446439733

[8] Park, E.S. Mechanical properties and antibacterial activity of peroxide‐cured silicone rubber foams Journal of Applied Polymer Science, 110 (2008) 3, S. 1723-1729, DOI: 10.1002/app.28750

[9] N.N. Fest- und Flüssigsiliconkautschuk – Ein Leitfaden für die Praxis Wacker AG, München, 2015

[10] Dunham, M.; Bailey, D.; Mixer, R.

New curing system for silicone rubber Industrial & Engineering Chemistry, 49 (1957) 9, S. 1373-1376, DOI: 10.1021/ie50573a029

© 2

017

Car

l Han

ser

Ver

lag,

Mün

chen

w

ww

.kun

stst

offte

ch.c

om

Nic

ht z

ur V

erw

endu

ng in

Intr

anet

- un

d In

tern

et-A

ngeb

oten

sow

ie e

lekt

roni

sche

n V

erte

ilern

.

http://www.sciencedirect.com/science/article/pii/S0142941802000818?via%3Dihub

http://pubs.acs.org/doi/abs/10.1021/ma0511200

https://link.springer.com/article/10.1007%2Fs40009-013-0150-2

http://onlinelibrary.wiley.com/doi/10.1002/pen.20698/abstract

http://www.hanser-elibrary.com/doi/book/10.3139/9783446439733

http://onlinelibrary.wiley.com/doi/10.1002/app.28750/abstract

http://pubs.acs.org/doi/abs/10.1021/ie50573a029



[11] Vennemann, N.; Bökamp, K.; Bröker, D.

Crosslink density of peroxide cured TPV Macromolecular Symposia, 245 (2006) 1, S. 641-650, DOI: 10.1002/masy.200651391

[12] Karrabi, M.; Mohammadian‐Gezaz, S.

Study of the cure characteristics and viscoelastic behavior of styrene‐butadiene rubber compounds by using a rubber process analyzer Journal of Vinyl and Additive Technology, 16 (2010) 3, S. 209-216, DOI: 10.1002/vnl.20228

[13] Dick, J. S.; Harmon, C.; Vare, A.

Quality assurance of natural rubber using the rubber process analyzer Polymer testing, 18 (1999) 5, S. 327-362. DOI: 10.1016/S0142-9418(98)00026-9

[14] Shibulal, G. S.; Naskar, K

Scrutinizing the influence of strain‐induced fiber orientation on the melt rheological behavior of short aramid fiber/thermoplastic elastomer composite using rubber process analyzer Journal of Applied Polymer Science, 128 (2013) 6, S. 4151-4163, DOI: 10.1002/app.38600

[15] Doğan, M.; Oral, D. D.; Yılmaz, B.; Savu, M.; et al.

Physical properties and cure characteristics of natural rubber/nanoclay composites with two different compatibilizers Journal of Applied Polymer Science, 121 (2011) 3, S. 1530-1535, DOI: 10.1002/app.33690

[16] Zhang, H.; Li, Y.; Shou, J. Q.; Zhang, Z. Y.; et al.

Effect of curing temperature on properties of semi-efficient vulcanized natural rubber Journal of Elastomers and Plastics, (2015). DOI: 10.1177/0095244315576243

[17] N.N. Product Data Sheet Silplus 60EX Momentive Performance Materials Inc., Leverkusen/Germany, 2016

[18] N.N. Product Data Sheet Akzo Nobel, USA, 2015.

[19] Shao, H.; Yu, Q.; Xiao, P.; Yao, W.; et al.

Study on viscoelastic properties of epoxidized trans-1, 4-polyisoprene by rubber process analyzer Polymer Science Series A, 54 (2012) 9, S. 760-766. DOI: 10.1134/S0965545X1208010X

© 2

017

Car

l Han

ser

Ver

lag,

Mün

chen

w

ww

.kun

stst

offte

ch.c

om

Nic

ht z

ur V

erw

endu

ng in

Intr

anet

- un

d In

tern

et-A

ngeb

oten

sow

ie e

lekt

roni

sche

n V

erte

ilern

.

http://onlinelibrary.wiley.com/doi/10.1002/masy.200651391/abstract

http://onlinelibrary.wiley.com/doi/10.1002/vnl.20228/abstract

http://www.sciencedirect.com/science/article/pii/S0142941898000269?via%3Dihub



http://journals.sagepub.com/doi/10.1177/0095244315576243

https://link.springer.com/article/10.1134%2FS0965545X1208010X



[20] Andersen, W. W.; Sezna, J. A.

Qualifying polymers using the RPA 2000 Rubber world, 232 (2005) 3

[21] N.N. D-MDR 3000 Application Manual

Montech Rubber Testing Instruments GmbH,

Buchen/Germany, 2008

[22] N.N. DIN 53529 part 3 Vulkametrie - Bestimmung des Vemetzungsverhaltens mit rotorlosen Vulkametern

[23] Svoboda, P.; Poongavalappil, S.; Theravalappil, R.; et al.

Effect of octene content on peroxide crosslinking of ethylene–octene copolymers. Polymer International, 62 (2013) 2, S. 184-189. DOI: 10.1002/pi.4277

[24] Maier, R.D.; Schiller, M.

Kunststoff Additive Handbuch

Carl Hanser Verlag, München, 2016.

DOI: 10.3139/9783446432918

[25] N.N. Crosslinking Elastomers and Thermoplastics AkzoNobel Functional Chemicals, USA, 2011

Bibliography DOI 10.3139/O999.02042017 Zeitschrift Kunststofftechnik / Journal of Plastics Technology 13 (2017) 4; page 233–253 © Carl Hanser Verlag GmbH & Co. KG ISSN 1864 – 2217

© 2

017

Car

l Han

ser

Ver

lag,

Mün

chen

w

ww

.kun

stst

offte

ch.c

om

Nic

ht z

ur V

erw

endu

ng in

Intr

anet

- un

d In

tern

et-A

ngeb

oten

sow

ie e

lekt

roni

sche

n V

erte

ilern

.

http://onlinelibrary.wiley.com/doi/10.1002/pi.4277/abstract

http://www.hanser-elibrary.com/doi/book/10.3139/9783446432918



Stichworte: Polydimethylsiloxan, Vernetzer, Rubber Process Analyzer, Vulkanisation, Silikonkautschuk Keywords: Polydimethylsiloxane, crosslinking agent, rubber process analyzer, curing, silicone rubber Autor / author: Fabian Verheyen M.Sc. Dr.-Ing. Ralf-Urs Giesen Prof. Dr.-Ing. Hans-Peter Heim Institut für Werkstofftechnik - Kunststofftechnik Universität Kassel Mönchebergstraße 3 34125 Kassel Deutschland

E-Mail: [email protected] Webseite: www.unipace.de Tel.: +49 (0)561/804-3670 Fax: +49 (0)561/804-3672

Herausgeber / Editors: Editor-in-Chief Prof. em. Dr.-Ing. Dr. h.c. Gottfried W. Ehrenstein Lehrstuhl für Kunststofftechnik Universität Erlangen-Nürnberg Am Weichselgarten 9 91058 Erlangen Deutschland Tel.: +49 (0)9131/85 - 29703 Fax: +49 (0)9131/85 - 29709 E-Mail: [email protected] Europa / Europe Prof. Dr.-Ing. Dietmar Drummer, verantwortlich Lehrstuhl für Kunststofftechnik Universität Erlangen-Nürnberg Am Weichselgarten 9 91058 Erlangen Deutschland Tel.: +49 (0)9131/85 - 29700 Fax: +49 (0)9131/85 - 29709 E-Mail: [email protected]

Amerika / The Americas Prof. Prof. hon. Dr. Tim A. Osswald, verantwortlich Polymer Engineering Center, Director University of Wisconsin-Madison 1513 University Avenue Madison, WI 53706 USA Tel.: +1 608/263 9538 Fax: +1 608/265 2316 E-Mail: [email protected]

Verlag / Publisher: Carl-Hanser-Verlag GmbH & Co. KG Wolfgang Beisler Geschäftsführer Kolbergerstraße 22 D-81679 München Tel.: +49 (0)89/99830-0 Fax: +49 (0)89/98480-9 E-Mail: [email protected]

Redaktion / Editorial Office: Dr.-Ing. Eva Bittmann Christopher Fischer, M.Sc. E-Mail: [email protected] Beirat / Advisory Board: Experten aus Forschung und Industrie, gelistet unter www.kunststofftech.com

© 2

017

Car

l Han

ser

Ver

lag,

Mün

chen

w

ww

.kun

stst

offte

ch.c

om

Nic

ht z

ur V

erw

endu

ng in

Intr

anet

- un

d In

tern

et-A

ngeb

oten

sow

ie e

lekt

roni

sche

n V

erte

ilern

.

zeitschrift kunststofftechnik journal of plastics technology · verheyen, giesen, heim curing...

Documents