CRP-208-287
CARBON-SILICA DUAL PHASE FILLER,
A NEW GENERATION REINFORCING AGENT FOR
RUBBER
PART III. ESCA AND IR CHARACTERIZATION OF
CARBON-SILICA DUAL PHASE FILLERS
Lawrence J. Murphy, Meng-Jiao Wang and Khaled Mahmud
Paper presented at a meeting of the Rubber Division, American Chemical Society.
Cleveland, Ohio, October 21-24, 1997
Paper No. 71
2
Carbon-Silica Dual Phase Filler,
A New Generation Reinforcing Agent for Rubber
Part III. ESCA and IR Characterization of
Carbon-Silica-Dual phase Fillers
Lawrence J. Murphy*, Meng-Jiao Wang, and Khaled Mahmud
Cabot Corporation, Billerica Technical Center
157 Concord Road, Billerica, Massachusetts 01821
Presented at a meeting of the
Rubber Division, American Chemical Society
Cleveland, Ohio
October 21-24, 1997
*Speaker
3
ABSTRACT
A series of carbon-silica dual phase fillers have been analyzed using Electron
Spectroscopy for Chemical Analysis and Infrared Spectroscopy. These techniques reveal
that the silicon in these materials is consistent with that in silica, and the predominant
carbon in the carbon phase is similar to the predominant carbon in carbon black. ESCA
and IR experiments demonstrate that the carbon-silica dual phase fillers are comprised of
composite aggregates containing silica and carbon phases, where the silica phase is
intimately distributed with the carbon phase. Quantitative ESCA reveals that these dual
phase fillers have an excess of oxygen over the expected stoichiometric amount for silica.
This excess oxygen is likely due to oxygen groups on the carbon phase Analysis of the
ESCA oxygen Auger lines in combination with various chemical treatments substantiate
the presence of carbon phase oxygen groups on the dual phase fillers. Quantitative
ESCA in combination with IR spectra of a dual phase filler after removal of the silica
phase, by HF treatment, suggests that the carbon phase of the dual phase filler contains
more oxygen groups than its carbon black counterpart.
4
INTRODUCTION
This is part three in a series of articles on the Carbon-Silica Dual Phase Fillers (CSDPF),
a new generation of materials for rubber reinforcement. The first two papers introduced
this new material, reviewing the reasons for its development and providing an overview
of its analytical and in-rubber performance properties.1,2 These new family of fillers are
produced by pyrolyzing petroleum based feedstocks and silicon based feedstocks, a
unique co-fuming technology developed by Cabot Corporation.1 Using this technology
it is possible to produce a family of new generation fillers ranging in surface area,
structure and silicon content.
In the first paper it was established that this new material was similar in morphology to
carbon black, comprised however, of individual composite aggregates containing carbon
and silica phases (a dual phase aggregate). The dual phase fillers were shown to have
higher polymer-filler interaction than a blend of silica and carbon black of equivalent
silicon content. The dual phase fillers were also shown to have lower filler-filler
interaction in comparison to either conventional carbon black or silica of comparable
surface area. As a result it was conjectured that the dual phase fillers would require less
coupling agent to bring about the same level of hysteresis as silica compounds. This was
subsequently confirmed.2 The second paper in this series reported on an in-depth study
of the dual phase fillers in relation to compounding design and mixing techniques for
application in passenger and truck tire tread compounds. It was revealed that with proper
5
mixing and compound formulations, the dual phase fillers can provide improved balance
of hysterisis with increased abrasion resistance when compared to carbon black and silica
compounds.
Among the many analytical techniques used in the first paper, describing the dual phase
fillers, ESCA (an acronym for Electron Spectroscopy for Chemical Analysis) also known
as X-ray Photoelectron Spectroscopy (XPS), and infrared spectroscopy (IR) were critical
to understanding the chemistry of the elements present. This paper will further describe
the results of these analytical techniques on the dual phase fillers. ESCA is a surface
chemical analysis technique that typically examines the top 7 nm of a solid material and
is considered an excellent qualitative tool, elucidating the presence of all elements
(excluding H and He) in concentrations greater than 0.1 to 0.2 mole % in the near surface
region of the material.3,4 Since this technique is useful in determining the chemical state
of elements, it was employed to determine the chemistry of the silicon, oxygen and
carbon present in the carbon-silica dual phase fillers.
Infrared spectroscopy is a well established analytical method. It has been used to study
the surface chemistry of many materials including silica5,6 and carbon black.7,8 Oxygen
functional groups on carbons have been investigated by a combination of IR,
neutralization reactions and chemical derivitization experiments.9-11 Infrared bands
characteristic of various oxygen functional groups have been summarized12
6
The purpose of this paper it to describe the chemical nature of the carbon, oxygen and
silicon in the dual phase fillers using ESCA and IR.
EXPERIMENTAL
ESCA
Two ESCA instruments, namely Hewlett-Packard 5950A and PHI 5500 ESCA
Spectrometers, were employed to obtain the spectra described herein. Samples were
ground and placed on double sided adhesive tape. Unless otherwise noted the ESCA
spectra are displayed as normalized intensities versus binding energies. The normalized
intensities were obtained by dividing the intensity of a given binding energy by the
maximum intensity in the spectrum. At times some of the spectra were offset for display
reasons. The binding energy scale of all spectra were corrected such that the Carbon (1s)
spectrum peak maximum was located at 284.6 eV.
Hewlett-Packard 5950A ESCA Spectrometer
This instrument used for the charging experiments, is equipped with an Aluminum Kα
anode (1486.6 eV) and a monochromator. The photoelectron take off angle was 38
degrees. An electron flood gun was used with the charging experiments, at times on and
at times off, as noted in the Figures.
PHI 5500 ESCA Spectrometer
7
This instrument employs a monochromatic Al Kα source equipped with an Argon
sputtering unit. The ESCA instrument used was equipped with flood gun capabilities and
used for all samples. The take off angle employed was 65 degrees. This instrument was
used to generate most of the ESCA data, except for the charging experiments.
Infrared Spectroscopy
In all cases the samples were run as KBr discs in a Nicolet Magna 550 FT-IR instrument
operated in the transmission mode. The concentration of the samples in the KBr were
held constant to 0.09% (w/w). The specimens were dried for two hours and the KBr for
four hours, both at 125 °C. 20 mg of carbon black and 980 mg of KBr were placed into
a stainless steel vial containing one steel ball and mixed for two minutes using a Wig-L-
Bug. A 45 mg aliquot of this mixture and 955 mg of KBr were placed into another
stainless steel vial containing one steel ball and mixed for two minutes on a Wig-L-Bug.
A 200 mg aliquot was placed into a die and evacuated (using a rotary pump) for three
minutes and then under vacuum pressed at 131 MPa for three minutes. Unless otherwise
stated all sample preparations were conducted in the laboratory, not in a dry box. The
specimens were placed into an IR holder and placed into the IR spectrometer. The
sample compartment was flushed with dry nitrogen for one hour. The IR scans were
obtained from 4000 to 400 cm-1 performing typically 100 scans at a resolution of 2 cm-1.
The spectra of the sample and KBr blank were corrected for increasing baseline to the
high wavenumbers such that after the correction, the baseline (absorbance versus
wavenumber) was flat. Then, half of the absorbance of the KBr blank spectra was
8
subtracted from the sample spectra. This procedure was found to adequately correct for
the KBr blank; subtractions using more than a factor of 0.5 resulted in negative
absorbances.
HF Treatment
Hydrofluoric acid (HF) treatment of the samples was carried out using 5% v/v
concentration of HF at ambient temperature for 10 minutes. After treatment, the samples
were washed with copious amounts of water and dried at 125 °C in preparation for
further analysis.
HNO3 Treatment
The nitric acid (HNO3) treatment, an oxidative treatment, of carbon black and dual phase
fillers was carried out in a small stainless steel heating drum. While the sample was
rolling 45 parts of (65%) acid per hundred of sample were slowly added. This mixture,
after a short “soak” period, was heated to 230 °C for 2 hours. After treatment the
samples were soxhlet extracted using water for 72 hours. The samples were then dried at
125 °C in preparation for further analysis.
RESULTS AND DISCUSSION
Electron Spectroscopy for Chemical Analysis and Infrared spectroscopy were used to
determine the chemical nature of carbon, oxygen and silicon elements as well as their
association with each other in the Carbon-Silica Dual Phase Fillers.
9
Electron Spectroscopy for Chemical Analysis
Survey scans from 0 to 1100 eV were obtained to qualitatively determine elements
present at or near the surface of these samples. Detailed spectra of the carbon(1s),
oxygen(1s), silicon(2p) and oxygen Auger region were obtained to determine the
chemical state of the elements. The analytical properties of the fillers used in this study
can be found in Table I.
Survey Scans
The survey scans reveal the presence of carbon, silicon, oxygen and sulfur on the dual
phase filler materials, and carbon, oxygen and sulfur on carbon black, see Figure 1. The
fumed silica sample, which is not shown, contained silicon and oxygen.
Carbon (1s) Spectra
The carbon (1s) peak binding energies and peak widths are summarized in Table II and
the spectra are given in Figure 2. The carbon peak binding energies for all samples are
located at 284.6 eV, because of the normalization procedure employed. The full width at
half of the peak height maximum (FWHM) are similar for the dual phase fillers and
carbon black. The silica sample on the other hand is broader. The more narrow carbon
(1s) line for carbon black and the dual phase fillers relative to silica suggests that the
carbon in the former two fillers is more graphitic in nature, as expected.13 The broader
10
carbon (1s) lines for the silica sample may also be explained by the insulating nature of
the silica.14 The dual phase fillers and carbon black contain a satellite peak at roughly
291 eV due to π-π* transitions. The presence of this satellite is indicative of aromatic
type carbon, similar to graphite.15 There is a noticeable lack of this satellite peak for the
silica samples, as expected because the carbon species on the silica sample are
adventitious carbon, not graphitic carbon.16
Two other regions useful in characterizing the graphitic nature of the carbon are the
plasmon loss band located approximately 30.5 eV upfield from the carbon (1s) line, or
about 315 eV, and the valance band region located between 0-30 eV. From the survey
scans the plasmon loss for the dual phase fillers and carbon black are found at
approximately 312 eV, again indicative of a graphitic like system, see Figure 1. The
valance band region is somewhat complicated by the oxygen (2s) peak at about 25 eV,
but the dual phase fillers and carbon black show evidence for the carbon valance band at
about 18 eV with a shape consistent with graphitic carbon.15 The information obtained
from the width of the C(1s) peak in combination with the presence of the carbon satellite
peak, plasmon loss peak and valance band suggests that the predominant carbon in the
dual phase filler is consistent with the graphitic like carbon in carbon black.
According to Briggs and Seah17 oxygen attached to carbon atoms induces shifts to higher
binding energies by 1.5 eV per C-O bond. This would suggest that for the carbon blacks
and dual phase fillers the species responsible for the peak maximum at 286.1 eV maybe
due to carbon single bonded to oxygen, at 287.6 eV to either carbon single bonded to two
11
oxygens or carbon double bonded to one oxygen, and at 289.1eV to carboxylate groups.
Similar binding energies were given by Donnet et.al.18 Although not illustrated here,
treating N234 carbon black and Carbon-Silica Dual Phase Filler A (CSDPF-A) with
nitric acid (oxidation) results in what appears to be a minor band near the 288 to 289 eV
region, indicative of carbon phase oxygen groups, such as carboxylate groups for
example. Attempting to deconvolute the carbon (1s) line, however, into various carbon
species maybe misleading due to the somewhat arbitrary assignments of position, shape
and baseline of the peak.
Silicon (2p) Spectra
As expected there is no silicon (2p) peak for the carbon black sample, see Table II and
Figure 3. The location of the silicon (2p) peak for the dual phase fillers (103.5 eV),
although slightly shifted to higher binding energies than the silica sample (103.2 eV), is
within the accepted range of silica.19 The silicon (2p) peak for the silica sample is
broader than the dual phase fillers, suggesting charging effects with the insulating silica
material, as previously discussed. There is no evidence for silicon carbide species as
evidenced by the lack of a peak at 100.4 eV.19 This is further supported by the fact
that the silicon content of CSDPF-A is reduced upon treatment with HF (silica is
dissolved by HF and silicon carbides are not), where only a very small silicon(2p) peak
located at 102.9 eV remains, almost undetectable from the noise. Although this binding
energy is low, it is not out of the range of silica,20 and not in the region for silicon
carbide. The concentration of the silicon species after HF treatment is 0.72 weight % Si
(ash method) and 0.6 weight % Si (ESCA analysis). The lower ESCA derived silicon
12
content compared to the bulk ash analysis suggests that the residual silicon is hidden
within the carbon matrix of the particles. This residual silica is probably in the particle,
unavailable to HF.
The Carbon-Silica Dual Phase fillers were determined not to be a blend of fumed silica
aggregates and carbon black aggregates. This was done by comparing the silicon (2p)
spectra of a blend of fumed silica and N234 to CSDPF-A with the flood gun turned on
and the flood gun turned off, a method developed by Barr.21 During the XPS analysis,
silica, because it is an insulator, will charge up in the spectrometer if the flood gun is off,
resulting in a shift to higher binding energies than when the flood gun is turned on. This
shift in binding energy for an insulator is the result of positive charge buildup on the
surface which acts as a drag on the outgoing photoelectrons resulting in a decrease in
kinetic energy and consequently an increase in the binding energy. The flood gun sprays
electrons onto the surface, neutralizing such charge buildup. Conducting materials, on
the other hand, do not experience this charge buildup and therefore if a conductor and
insulator are present in a sample, differential charging may result. Barr has found that if
an insulator and conductor are well dispersed, then the conductive nature of the sample
will be determined by the property of the material in higher concentration. Carbon black
is sufficiently conductive such that it does not charge up in the spectrometer. If the
Carbon-Silica Dual Phase fillers were a simple blend of fumed silica and carbon black,
then the silicon (2p) peak position would be affected by the flood gun being turned on
and off. As can be seen from Figure 4, when the flood gun is off the silicon(2p) peak for
13
the blend moves to 112.9 eV (FWHM of 2.2 eV) from 103.2 eV (FWHM of 1.5 eV)
when the flood gun is on. There is also an increase in peak width when the flood gun is
off, consistent with non-uniform charging within the sample.22 In contrast, the peak for
the CSDPF-A remains relatively unaffected by the flood gun experiment, where the peak
is located at 103.4 eV with the gun off and at 103.5 eV with the gun on. The fact that the
dual phase filler is essentially unaffected by the actions of the flood gun supports the
claim1 that the CSDPF-A is not a mixture of silica aggregates and carbon black
aggregates, rather the silica is an integral and finely dispersed part of the dual phase
fillers.
The slight shift in binding energy of the dual phase fillers relative to silica may be
explained by a number of possibilities such as: small silica domains in the dual phase
filler relative to silica producing cluster shifts,23 charging errors affecting the binding
energies24 or a modest change in the bonding environment of the silicon. Barr25 has
argued that the presence of C-O bonds associated with silicon, which are more covalent
than Si-O bonds, should enhance the ionicity of the Si-O bonds resulting in a Si(2p) shift
to higher binding energy relative to silica. It has also been illustrated that silanols can
affect the Si(2p) peak position, shifting it to higher binding energies.26
The atomic concentrations determined by ESCA maybe found in Table III. Conversion
of the silicon atomic percentages to weight percentages for the dual phase fillers -A, -E
and -F results in 6.4%, 7.4% and 4.8%, respectively. If the dual phase material had a
homogenous distribution into the core of the particles, then one would expect the ESCA
14
and ash data to agree. If an element is more concentrated in the core of the particle
(probably depleted at the surface), then one would expect the ESCA data to be lower than
the ash data. Comparison of the ESCA determined weight percents with the bulk
composition (Table I) for the three dual phase fillers suggests that the silicon is
apparently slightly more concentrated near the surface of the particles than in the core.
This is consistent with the model presented in part I of this series of articles.1
Oxygen (1s) Spectra
As can be seen from Figure 5 and Table II the dual phase fillers and silica sample have
peak binding energies in the 532 to 533 eV region, which is in the vicinity expected for
silica.27 As in the case of the silicon (2p) spectra, the oxygen (1s) peaks of the dual phase
fillers are shifted to slightly higher binding energies (532.7 eV) than the silica sample
(532.5), which may support the hypothesis of a higher silanol to silicon ratio in the dual
phase fillers than for silica, as seen by ESCA.28 The peak width for silica is larger than
for the dual phase filler, again probably a result of charging effects with the silica sample.
The O(1s) of carbon black also falls within the same binding energy range, which is
expected based on work done by Barr15 and Darmstadt.29 Barr used the following
binding energy assignments to estimate the presence of various functional groups: C-O-C
(B.E.≈ 532.4 eV), C=O (B.E.≈ 531.8 eV), O-C=O (B.E.≈ 533.3 eV) and C-OH (B.E.≈
532.8 eV). Clearly these binding energies fall under the peak envelope of the N234,
where the peak maximum is consistent with the C-OH group.
15
A charging experiment was conducted, similar to that as for the silicon (2p) spectra,
where oxygen(1s) spectra of a blend of fumed silica and N234 was compared to CSDPF-
A by obtaining the spectra with the flood gun turned on and turned off. As can be seen
in the Figure 6 the peak binding energy for the CSDPF-A is unaffected by the flood gun
experimen