Title:

A uthor(s):

Submitted to:

Los Alamos - - - ~~~~ - -

N A T I O N A L L A B O R A T O R Y

THE STRUCTURE OF CARBONBLACK AND ITS COMPOSITES WITH ELASTOMERS: A STUDY USING NEUTRON SCA’ITERING

JUL I9 ’i936

R P. Hjelm, W. Wampler, M. Gerspache PSI-I

Invited talwpaper to be presented at “Mechanism of Reinforcement and Rubber Filler Interaction,” Hannover, Germany, 27-28 Jun 96

Los Alamos National Laboratory, an affirmative action/equal opportunity employer, is operated by the Univerjity of California for the US. Depattment of Energy under contract W-7401ENG-36. By acceptance of this article, the publisher recognizes that the US. Government retains a nonexclusive, royalty-free license to publish or reproduce the published form of this contribution, or to allow others to do so, for U.S. Government purposes. The Los Alamos National Laboratory requests that the publisher identify this article as work performed under the,auspices of the US. De R F o g k X 3 6 6

DISTRIBUTION OF MIS DOCUMENT IS U N L I W O N . - -_ ~ ~ -~ -~ __-.~ ~ - _ _ .I______ ~ . . ~.. ~- ~ ~

DISCLAIMER

Portions of this document may be illegible in electronic image products. Images are produced from the best available original document.

The Structure of Carbon Black and its Composites with Elastomers: A Study using Neutron Scattering

Rex P. Hjelm MLNSC, Los Alamos National Laboratory, Los Alamos, NM 87545

Wesley Wampler and Michel Gerspacher Sid Richardson Carbon Company, Fort Worth, TX 76106.

ABSTRACT We have been exploring the use of small-angle neutron scattering and the method of contrast varia-

tion to give a new look at very old problem-reinforcement of elastomers by carbon black in durable rubber products. The method has yielded some interesting information on the structure of an experimen- tal carbon black, HSA, and on the associations of HSA in polyisoprene composites.

Carbon black has a hierarchy of structures consisting of particles covalently bound into aggregates, which in turn associate by weak interactions into agglomerates. We found that in HSA the aggregates are rodlike, containing an average of 4-6 particles. The aggregates have an outer graphitic shell and an inner core of lower density carbon. The core is continuous throughout the carbon black aggregate. Con- trast variation of swollen HSA-polyisoprene gels shows that the HSA is completely embedded in poly- isoprene and that the agglomerates are formed predominantly by end on associations of the rodlike a

Further studies on production carbon blacks suggest that the shell-like aggregate structure is present in commercial carbon blacks.

gregates. The surface structure of the carbon black appears smooth over length scales above about 10 R

INTRODUCTION Carbon black (CB)-elastomer composites are examples of “classical systems” of heterogeneous mi-

crophases of matrix and reinforcement. The carbon black consists of a hierarchy of structures in which spheroid particles are fused into aggregates which associate by van der Waai‘s forces to form agglomer- ates. There is a substantial body of work on CB structures and CB-elastomer composites. Even so, over 90 years after the discovery of the effect of CB on rubber, there is controversy on what the reinforcing mechanism is. Ideas center on particle size and surface structure relating to associations with the elas- tomer, aggregate shape and the extent and morphology of agglomerates. In the interactions with the elastomer, the strength of the CB-polymer binding and polymer chain configuration and entanglements are considered highly important. A clearer understanding of the mechanical properties of reinforced rub- ber requires a better picture of the structure of the material and its components in situ.

Our objective is to provide structural information on CB and CB-elastomer composites. This in- cludes the morphology and internal structm of CB particles and aggregates; the association of aggre- gates in rubber composites; the associations of elastomer and carbon blacks. We also need to show the generality of the structural feature in different carbon blacks having different mechanical- properties when composited with rubber. In this paper we discuss the structure of aggregates as probed by neutrons.

SMALLANGLE NEUTRON SCATTJ3RZNG

Table I Scattering Lengths of Some Light Elements and Hydrogen Isotopes

Element Scattering length (fm)

‘H -3.7403 zD 6.675 C 6.6484 N 9.2600 0 5.8050

Small-angle neutron scattering (SANS) is an important technique in the battery of methodologies for structural charac- terization. The basic physical concept behind neutron scatter- ing is the same as for electrons or x-rays, but for neutrons the scattering is produced almost entirely by interactions with the nuclei. These interactions are very small relative to that of an electron or an x-ray, consequently the neutrons are relatively more penetrable. Because of the penetrability of the neutron SANS provides nanoscale and molecular scale structural in- formation on materials in the bulk. Thus, one can study car- bon black suspended in fluids or in situ with elastomers or in

a

2

containment apparati needed for special environments. An important aspect of the scattering interaction of neutrons with nuclei is that the scattering length, the measure of scattering amplitude, is not mono- tonic with atomic number (Table I), unlike radiation that interacts primarily with electrons, such as electrons or x-rays. Consequently, neutrons are able to see the the distribution of light elements. Fur- ther, there are often significant differences in the scattering from isotopes of the same elements. The difference between hydrogen and deuterium (Table I) is particularly important in this regard.

Sample structure as seen in -the SANS experiment is fluctuations in the scattering length density, ‘ p(r), as a function of position in the sample, r. The scattering length density at r is the sum of the scattering lengths of the atoms (see Table I) in the volume including r divided by that volume. Thus, p(r) reflects microscale structure in the sample in both density and chemical composition. The scat- tered intensity, I(Q), is measured as the absolute differential cross section per unit scattering mass (cm2 mg-’) as a function of the magnitude of the scattering vector, Q. For elastic scattering events, Q = (4d) sine, where X is the neutron wavelength and 8 is half the scattering angle. The intensity of scat- tering from p(r) is given by the Fourier transform relationship. Though this relationship can be stated in general form, we will discuss. it for the case of a system containing a collection of particles uncorre- lated in position and orientation, modifying it later for the more general case. The relationship between the structure and scattering intensity is,

,

where o represents a spherical average. The expression in the angle brackets is the spherical averaged particle form factor, <P(Q)>. The constant, K, in ~ q . (1) is itself proportional to L\P*, = P, - ps , which is the contrast between the average scattering length density of the particle, c, and the solvent scattering length density, p,. Thus, I(Q) gives info&on about particle size and shape. Because the wavelength of neutrons used in the SANS measurement is of order 1 to 10 A, the values of Q over which the measurements are made probe length scales on the order of a few to several hundreds of Angstroms (length scales - Q-I).

PARTICLE STRUCTURE CHARACTERIZATION The size and scattering mass of the scattering body can be determined by Guinier analysis, which

applies over the domain Q < K’, where R is the average radius of the particles. The Guinier approxi- mation has the form :

where 9 is the volume fraction of scatterers of volume Vp, M is the mass of the scatterers and & is the radius of gyration.

fitting the observed scattering, over a wide rauge of Q, to a particle form factor. Once a general idea of the particle size and shape is obtained, the models may be further refined by

Over length scales such that QR >> 1 the scattering can be described as a power-law :

I(Q) = Z,Q-=. (3)

Here, (r and I,, arc constants, with the value of a being indicative of the surface “smoothness”. For the special case of smooth interfaces, a = 4 and the scattering is termed Porod-Scattering. In this case, Z, = 27&4 A@, where A is the interfacial area per unit volume of the sample.

In the case where there are particle interaction, so that the positions and or orientations are corre- lated in some way, it is necessary to reinterpret Eq. (1) as

In this expressions S’(Q) is the structure factor. S’(Q) arises from particle correlations in solution and can be formulated to take into account the possibility that the interacting particles are not spherical and that they might by polydisperse in size and shape.

3

5 -

4 - n

o! u

z.f! )-r

a . 3 -

2 -

1 -

0 -

-1

electron microscopy (“EM). The lack of - spatial information in the SANS experi-

ment is a result of the loss of phase infor- . mation in the measured I@). We have - outlined a means of recovering spatial in- - 7 formation in composite materials by the

method of contrast variation. 23 In the method of contrast variation, which was developed for the study of biological mac- - romolecules and assemblies in solution 4,

we change Ap2 by suspending the sample - in fluids having different mixtures of deu- : terated and protonated solvent (Fig. 1). - This method can be used to contrast out

one or another of the components in a - complex mixture, or it can be used to de- - termine the internal structure of one, say, : CB. We implemented this approach in - - studies of CB and CB-elastomer compos- : ites. The CB-elastomer composites were

prepared as “bound” rubber 3, which re- mains after extensive high temperature ex- traction with good solvents. This material

A

4 : si02 (Luiox)

Qsi P G

’I Y-

V Polyisopm

f 0 - Q . J , -

For a homogeneous solution of non-interacting particles the scattering has a well defined behavior

( 5 )

When Ap = 0, this is the contrast match point The functionsZn(Q), Z,(Q), and Z,(Q) are the basic scattering functions. The f i t two arise, respectively, from the solvent-excluding parts of the structure, a(r), the scattering from the internal scattering length density fluctuations, c(r). The last function is the scattering due to correlations between Q(r) and c(r). In this representation the structure is defined as the s u m of the shape and internal structure terms; thus, p(r) = ps+ ApQ(r) + &). Eq. (5) is an ap- proximation for heterogeneous systems, such as studied here, as the different terms are better repre- sented as sums over the different contrasts present in the sample.

The analytical methods outlined above (Eqs. (2)-(4)) can be applied to the method of contrast varia- tion. In particular the Guinier analysis yields,

as a function of contrast given by

ZtQ) = AP2Za(Q) + APZw(Q) + Z,+Q)

for the contrast dependence at Q = 0, and

4

for the contrast dependence of 4. In Eq. (7) R, is the radius of gyration of Q(r), a is proportional to the second moment of c(r) and is proportional to the square of the first moment of ((r). Thus the contrast variation method provides five structural parameters based on the Guinier analysis: Volume and 5,. from the changes that occur in I(0) with ps, and the three parameters in Q. (7).

HSA-Polyisoprene X HSA '-E - I

0.01 0.1 I 0 Q (A9

Figure 2: SANS of HSA-polyisoprene Composite Geb and HSA Suspended in different Fractions of Deu- terocyclohexnne Data: x, HSA, H, HSA-plyisoprene composite gel.

CARBON BLACK STRUCTURE IN SUSPENSION AND WITH ELASTOMER For our studies we used the experimental carbon black, HSA. There are substantial changes in the

Table 11 Analysis of the of the contrast- dependent,Guinier analysis using Eqs. (6) and (7) for the HSA sample is summa-

V, (nm3) P, (cm-') & (nm) a B rized in Table XI. The large value of a is highly significant, for it shows that the

5'4 (3) lo' 5*7 (Os3) lolo 29.3 (4) 2.0 (3) lo-' particles have a shell-core structure, con- sistent with TEM ' and scanning tunneling

microscopy 6. with an outer shell with density like that of graphite. The core of the particle has density like that of amorphous carbon, but with voids. Most significantly, the particles are fused together by the amorphous cores to form the aggregate. I&) determined using Eq. (5) was fitted to a prolate ellipsoid of revolution showing that the aggregates are approximately 290 8, by 1500 8, and are rodlike, having lit- tle branching, with 5 to 6 particles on average. This result is consistent with stereo TEM studies. ' Fi- nally, I,(Q) obeys Porod's Law (Eq. (3)); thus, the particles have smooth surfaces on length scales

neutron scattering of HSA and HSA-polyisoprene (HSA-PI) gels with, fc642 (Fig. 2):

Structural Parameters for HSA

4' 5

Figure 3: Schematic of the Carbon Black Aggregate Structure: The structure consists of a linear amy of spheroid particles. The average aggregation number is 4 to 6 with particle sizes being in the range of 240 to 290 A. The outer shell of the structure (black) consists of graphitic carbon crystallites, with a core (gray) of less dense void-filled (white) carbon.

0.1

Figure 4: Comparison of Scattering from HSA suspen- sions and Gels of Bound Rubber Computed for fcSDl2 = 0.07: Data: 0, HSA, 0, HSA-polyisoprene.

shown in Fig. 5. Some features of S'(Q) probably arise

greater than 10 to 20 A. These conclusions are summarized in schematic form in Fig. 3. Studies on the production CB's, N330 and XLH81 suggest that the shell-core structure found in HSA is present in these CB's, as well.

These conclusions have important implica- tions on the mechanism of reinforcement of HSA. Ideas about rubber composite properties must take into account the short, rigid rodlike character of the carbon black aggregates. Fur- ther, the mechanism of polymer binding must take into account the upper limit of the length scale for CB surface roughness determined from the SANS studies.

The SANS of HSA-polyisoprene bound rubber gels is significantly different from that of HSA alone (Fig. 2). These data, too, can be fit using Eq. (5). When Q is less than ap- proximately 10 k', the fits suggests a mini- mum at a Ap near the;,, value for HSA (Table n)? For Q greater than 10 there is a shift in the Ap for the minimum scattering intensity towards the computed; of 3.2 10" cme2 of the composite, implying that the con- trast for CB is from the solvent-polyisoprene mixture. Thus, CB is almost completely coated with elastomer.

The scattering at fCdD,* = 0.07, where only CB scattering should be observed, can be calculated from the data in Fig. 2 by interpola- tion using Eq. (5). The result of this calcula- tion is shown in Fig. 4. For Q < 0.02 A-' the scattering from HSA is considerably greater than that from the HSA-polyisoprene compos- ite. For Q between 0.02 and 0.07 the HSA intensity is slightly less than that of the composite. The scattering from the two sam- ples becomes indistinguishable for Q greater than 0.07 A-'. We have shown that the scat- tering from HSA alone in suspension (Fig. 2) is very close to that expected from non- interacting particles, and when fc6D12 is be- tween 0 and 0.25 scattering reflects the carbon black aggregate shape with little contribution from the internal structure. We, thus use Eq. (4) to calculate S'(Q), the result of which is

from the fact that the CB aggregates are elon- gated. That s'(Q) is significantly smaller-than unity at low Q suggests that there is strong exclusion of CB particle neighbors in the HSA-PI composite. It is likely then that on average each CB aggregate is separated by a considerable amount of polymer that prevents the aggregate from making lateral associa- tions. On the other hand the amount of CB in this sample, 65% by weight in the dried material, is above the percolation lit. Thus, the CB aggregates must be touching, in which case these results show end on association of the rodlie aggregates.

I

1

? 0.5 v1

6

When the incoherent backgrounds are accounted for, the scattering in the calcu- lated at fcbD,2= 0.07 (Fig. 4) falls off as Porod‘s law (Eq. (3)) for Q -c 0.02 indi- cating that the smooth surfaces are also pres- ent in HSA in the solvent-impregnated poly- isoprene. The scattering in this same Q domain from samples near to the HSA con- trast match point fall of with power laws in Q (Eq. (3)) intermediate between -4 and -3. This observation suggests that the surface of the contrasted particles appear rough at high polyisoprene contrasts. This can be inter- preted as beiig from interpenetrating solvent and polyisoprene-rich phases that surround the HSA-PI aggregates.

I I ACKNOWLEDGMENTS This work was supported by the Office of

Basic Energy Sciences of the Department of Energy. This work benefited from the use of the Low-Q Diffractometer at the Manuel La- jan Neutron Scattering Center of the h s Alamos National Laboratory which is s u p

sci- Of En-

ergy under contract W-7405-ENG-36 to the University of California

0 . 1 p , , , , , , . * , * * j 0.01 o. 1

Q (K*)

Figure 5: Structure factorfrom the scattering of HSA in ported by the Office Of Basic the Composite Gel of Bound Rubber: Calculated values: ences Of the united 0.

REFERENCES 1. Hess and C.R Hurd, “Microstructure, morphology and general physical properties”, Q&QU

H..&, Dekker, New York, 1993, pp 91-106, and references there in. 2 R.P. Hjelm, W. Wampler, P.A. Seeger, and M. Gerspacher, “The microstructure of carbon black

A study using small-angle neutron scattering”, J. Mat. Res., 9,3210-3222 (1994) 3. R.P. Hjelm, W. Wampler, and M. Gerspacher., “Variation of solvent scattering-length density in

small-angle neutron scattering as a means of determining structure of composite materials”, Md. Res. SOC. Symp. Proc., 376,303-307 (1996).

4. K. Ibel , and H.B. Stuhrma~, “Comparison of neutron and x-ray scattering of dilute myoglobin solutions”, J. Mol. Biol., 93,255-265 (1975).

5. T.C. Gruber, T.W. Zerda, and M. Gerspacher, “Three-dimensional morphology of carbon black aggregates”, Carbon, 31, 1209-1210 (1993).

6. J.-B. Donnet and E. Custodtk, “Ordered structures observed by scanning tunpeling microscopy at atomic scale on carbon black surfaces”, Carbon, 30, 813-815 (1992).

7. T.W. Zerda, H. Yang, and M. Gerspacher, “Fractal dimension of carbon-black particles”, Rub- ber Chem & Tech., 65,130-136 (1992).

DISCLAIMER

This report was prepared as an account of work sponsored by an agency of the United States Government. Neither the United States Government nor any agency thereof, nor any of their employees, makes any warranty, express or implied, or assumes any legal liability or responsi- bility for the accuracy, completeness, or usefulness of any information, apparatus, product, or process disclosed, or represents that its use would not infringe privately owned rights. Refer- ence herein to any specific commercial product, process, or service by trade name, trademark, manufacturer, or otherwise does not necessarily constitute or imply its endorsement, recom- mendation, or favoring by the United States Government or any agency thereof. The views and opinions of authors expressed herein do not necessarily state or reflect those of the United States Government or any agency thereof.