10

Click here to load reader

XPS study of the halogenation of carbon black—Part 2. Chlorination

Embed Size (px)

Citation preview

Page 1: XPS study of the halogenation of carbon black—Part 2. Chlorination

Pergamon

0008-6223(94)00111-l

Carbon, Vol. 33, No. I, pp. 63-72, 1995 Copyright 0 1995 Elsevier Science Ltd

Printed in Great Britain. All rights reserved 0008.6223/95 $9.50 + .OO

XPS STUDY OF THE HALOGENATION OF CARBON BLACK - PART 2. CHLORINATION

EUGENE PAPIRER,* RENAUD LACROIX, JEAN-BAPTISTE DONNET,

GERARD NAN& and PHILIPPE FIOUX Centre de Recherches sur la Physico-Chimie des Surfaces Solides,

24, Avenue du President Kennedy, 68200 Mulhouse, France

(Received 6 April 1994; accepted in revised form 13 September 1994)

Abstract-Three furnace carbon blacks, a thermal black, and an electrically conductive sample were sub- mitted as received to chlorination, at 45O”C, with a mix of chlorine and carbon tetrachloride vapors. The treated samples were examined using chemical and spectroscopic (XPS) methods. The interpretation of the XPS spectra is facilitated by comparison with spectra of reference materials. XPS detects various types of carbon-chloride bonds formed essentially by addition or hydrogen substitution processes. The amount of fixed chlorine is dependent on the origin of the blacks, and there is no direct relationship between spe- cific surface area and chlorine uptake. It is shown that chlorination is not limited to the external surface, nor is it to a sub-superficial layer. In fact, it proceeds inside the carbon black particles to a depth depend- ing on their origin.

Key Words-Carbon blacks, chlorination, XPS, 2,4,5,6 tetrachloro m-xylene, 1,2,3,4 tetrachloronaph- thalene-bis hexachlorocyclopentadiene adduct, chlorination mechanism.

1. INTRODUCTION

The amount of work devoted to the chlorine/graphite system is less important than that on the bromine/ graphite system. The fixation of chlorine on natural or synthetic graphites has been studied, in particular, by Hennig[ I] and Insa et al. [2,3]. From the X-ray dif- fraction spectra of the graphite/liquid chlorine system, determined at low temperature, Insa and Seidel[3] demonstrated that chlorine, like bromine, forms an intercalation compound tending to the ultimate com- position C&l. However, the intercalation of chlorine is only possible at low temperature (<-30°C). At room temperature, the intercalation is negligible[4]. Moreover, as with bromine, that part of the chlorine fixed at low temperature is not desorbed when the sample is brought into contact with air, at room tem- perature. According to Hooley[S], the chlorine con- tent of the “residual compound” is lower than that of the corresponding “residual” graphite/bromine com- pound. Probably, at low temperature, part of the chlo- rine is irreversibly fixed on the peripheral surface of the graphitic structure in the form of covalent bonds of type C(sp3)-Cl,. In that way, the departure of in- tercalated chlorine is hindered when the sample is left at atmospheric conditions[6].

Concerning carbon blacks, in a recent review, Ban- sal and Donnet[7] describe the formation of carbon- halogen surface compounds when heating carbon blacks with different halogenation reagents, under var- ious experimental conditions. They noted that the amount of fixed halogen depends upon the origin of the carbon and its oxygen and hydrogen contents. Boehm et al. [8] and Puri et al. [g-10] have shown that

*To whom correspondence should be addressed.

Cl2 is strongly and irreversibly bound to various char- coals and carbon blacks at elevated temperatures (preferentially between 400-500°C). Fixed chlorine can only be partially eliminated upon evacuation at high temperature (12OO’C) or upon boiling in concen- trated alkaline hydroxide solution [ 111.

For the chemical bonding, several mechanisms are operative: addition to unsaturated ethylenic sites[9,12- 191, exchange with chemisorbed hydrogen[9-11,13-161 and strong adsorption due to formation of charge trans- fer complexes associated with the electron acceptor character of C1,[21], simple physical adsorption[l8,20], entrapment in very narrow pores[ 161, and surface ox- idation (from aqueous solution) where water is chem- ically involved.

In a previous paper, we[22] reported the results of the bromination of carbon blacks using either liquid or gaseous bromination conditions. We compared es- sentially the results of elemental analysis and electron spectroscopy (XPS). The interpretation of the XPS re- sults, both qualitatively and quantitatively, allowed us to assess the fixation mechanisms of bromine and, moreover, to demonstrate that the fixation of bromine is not limited to the surface or to a surface layer, but it may well proceed into the bulk of carbon black de- pending on its origin (furnace or thermal). It is the object of this paper to complete these results by ex- amining, comparatively, the fixation of chlorine on the same carbon surfaces.

2. EXPERIMENTAL

2.1 Carbon blacks The general characteristics of the 5 carbon blacks

studied (a conductive black: XP-2, three furnace blacks:

63

Page 2: XPS study of the halogenation of carbon black—Part 2. Chlorination

64 E. PAPIRER et al.

N 115, N 326, N 772, and a thermal black: MT N 990, all from Degussa) have been presented in the previous paper[22].

2.2 Method of chlorination The selected chlorination method consists, firstly,

in an outgassing of the carbon black, at 45O”C, un- der a nitrogen flow. Then, a flow of gaseous chlorine and nitrogen, which has percolated through carbon tetrachloride, is contacted with carbon black. Finally, the reaction mix is allowed to cool under a nitrogen blanket. The samples are stored under atmospheric conditions. This procedure leads to reproducible results.

2.3 XPS analysis The same procedure, as earlier[22], using a Leybold

Heraeus spectrometer (Mg Kcz source), was applied to chlorinated carbon blacks.

For fitting of the XPS spectra and assigning of the components, we referred to the recent Database worked out by Beamson and Briggs[23] for chlorinated poly- mers and to two model compounds: 2,4,5,6 tetrachloro m-xylene (T.C.M.X -Aldrich) and 1,4: 5,8dimethano-

tripheny1ene,1,2,3,4,5,6,7,8,9,10,11,12,13,13,14,14- hexadecachloro-1,4,4a,4b,5,8,8a, 12b-octahydro, ob- tained (from Aldrich) by a Diels Alder condensation procedure of 1,2,3,4 tetrachloronaphthalene and hexa- chlorocyclopentadiene, denoted in short in this study T.N.H.C.P., which we were the first to examine by electron spectroscopy. The corresponding spectra are represented in Fig. 1.

2.3.1 CI2ppeak. Three Cl forms can be differ- entiated from the structural formula of T.N.H.C.P: Cl bonded to sp2 carbons of the aromatic ring and to olefinic carbons (noted as Cl(l) in the formula), Cl bonded to sp3 carbons of the five-membered cycle (noted as Cl(2) in the formula), and Cl belonging to CClz of the five-membered cycle (noted as Cl(3) in the formula). These three forms are not differentiated on the C12p peak, but contribute to the component (1) having a binding energy Eb = 200.6 ? 0.2 eV (taking as a reference the binding energy of the contamination carbon, i.e., 285.2 i 0.1 eV). The component (2), of very low intensity, is shifted 1.7 rt 0.2 eV from (1) in the direction of decreasing binding energies, and cor- responds to traces of chlorides. This component is, however, very apparent in the spectra of T.C.M.X, and is attributed to FeCl, impurities. (FeCI, is used for the synthesis of T.C.M.X). The component (3), which is very weak, is shifted 3.0 + 0.2 eV from (I), and possibly originates from traces of HCI stemming from the hydrolysis of FeCI,.

2.3.2 Cls peak. Two main components are noted on the spectra of T.C.M.X.: One (1) at a bond- ing energy of 285.2 k 0.2 eV and the second (2) shifted 1.10 k 0.2 eV (1) in the direction of increasing bind- ing energies.

Component (1) originates from three types of car- bon bonds having very close binding energies: Csp3 from methyl groups, Csp2 from C carrying those groups, and Csp3 from hydrocarbon contamination.

The peak area corresponding to the first two types is similar to the area of component (2) corresponding to Csp2 of chlorinated carbon atoms. Since the total area of component (1) is more than twice the area of com- ponent (2), it is concluded that hydrocarbon contam- ination represents about 50% of component (1).

Component (2) corresponds to the sp3 carbons co- valently bonded to chlorine. From the area of the component (1) of the C12p peak, one may estimate the theoretical area of the component Cls (2): Area Cls (2) = (20/58) x Area C12p. The atomic sensitivity fac- tors for Cls and C12p are, respectively, 0.20 and 0.58 (Leybold DSlOO Database). We find Area Cls (2) = 790 * 10% cps eV, compared to the actual area of component (2), which is equal to 781 t 10% cps eV (i.e., a value in close agreement with the theoretical one).

Components (3), (4) and (5), of low intensities, are shifted 1.80 ? 0.2 eV, 3.00 & 0.2 eV, and 4.20 k 0.2 eV, respectively, from (1) in the direction of increas- ing binding energies. They are attributed to C partic- ipating in carbon-oxygen groups: -C-OR, ethers (3), C=O (4), and carboxyl groups (5).

The Cls peak of T.N.H.C.P. is more complex. Be- sides component (1) corresponding to nonchlorinated carbons of the cycle (C(1) on the formula) and, to a lesser extent, to aliphatic carbons of the hydrocarbon contaminants, three other components corresponding to carbons linked to Cl are evidenced. Component (2), shifted 0.85 f 0.2 eV from (1) in the direction of in- creasing binding energies, can only be monochlori- nated sp2 carbons of the aromatic ring (C(2) in the formula). Component (3), 0.5 * 0.2 eV away from (2), corresponds to sp3 carbons of the five-membered cycle (C(3) in the formula). Component (4), shifted 1.5 +- 0.2 eV from (2), corresponds to dichlorinated carbon atoms of the five-membered cycle (C(4) in the formula).

From the peak C12p, one may calculate the theo- retical area of the three components:

Area Cls (2) = l/2 [(20/58) x Area C12p],

Area Cls (3) = l/4 [(20/_58) x Area C12p],

Area Cls (4) = l/8 [(20/58) x Area C12p].

The calculated area A Cls (2) is 1456 + 10% cps eV, whereas the measured one is 1480 + 10% cps eV on the fitted spectra. A Cls (3) = 728 * 10% cps eV compared to 740 * 10% cps eV for the fitted peak. Fi- nally, A Cls (4) = 364 k 10% cps eV (i.e., a value very close to the area measured on the fitted peak, which is equal to 370 + 10% cps eV). The good agreement between calculated and measured (fitted) areas gives much support to our fitting procedure.

In short, Cl that are covalently bonded in mono- or dichlorinated groups cannot be differentiated from the CI2p peak. The binding energies of the corre- sponding photoelectrons (C12p 3/2) are of the same order of magnitude: 200.55 ? 0.1 eV in the chlorinated polymers[23], taking as a reference the carbon of con-

Page 3: XPS study of the halogenation of carbon black—Part 2. Chlorination

XPS study of the halogenation of carbon black-Part 2 65

1400

1200

1000

600

600

400

200

0

i5oc

loot

50t

I

1 CH3

(a) Carbon 1 s

T.C.M.X. c’ A

T.N.H.C.P-

, 4 I,

960 965 970

1 T.N.H.C.P. 1

WCI

(WI

(2) Cl

Cl (1)

Cl (1)

Cl)CI

too0 1 1500-

1200-

800-

400.

0,

4000

3000

2000

1000

a

Cl (1)

(b) Chlorine 2~

T.C.M.X. t

‘-k--2 T.N.H.C.P.

1045 1050 Kin. energy [ eV ]

Fig. 1. Fitted C12p peak and Cls peak of reference products: 2,4,S,6-tetrachloro m-xylene (T.C.M.X.) and 1,2,3,4-tetrachloronaphthalene-bis hexachlorocyclopentadiene adduct (T.N.H.C.P.).

taminants having a binding energy of 285.0 eV. How- ductive effect of CI on the p carbon atom of the aro- ever, the chemical shifts induced on the carbon atom matic cycle (sp2) is negligible. The chemical shift core levels differ significantly and, hence, the compo- induced on sp3 carbons, when covalently bonded to

nents may be identified. When a hydrogen atom of an Cl, is larger. In the case of T.N.H.C.P., the deviation sp2 carbon is substituted by Cl, the binding energy of we observed between the binding energies of Csp2 and Cls electrons increases by 1.00 k 0.2 eV[23]. The in- Csp3 bonded to one Cl atom is about 0.5 Z?Z 0.1 eV. In

Page 4: XPS study of the halogenation of carbon black—Part 2. Chlorination

66 E. PAPIRER et al.

polymers, according to Beamson and Briggs[23], it is about 1 .O t 0.2 eV. For chlorinated carbon blacks, by fitting of the Cls peak and making digital spectra sub- traction, we noticed that the best adjustment was ob- tained when taking for the deviation between the components corresponding to monochlorinated Csp2 and Csp3 a value equal to 0.85 f 0.1 eV. In the case of T.N.H.C.P., the binding energies of monochlori- nated sp3 carbons and dichlorinated carbons of the five-membered cycle differ by 1 .O + 0.1 eV. For chlo- rinated polymers, the deviation is 1.5 rf- 0.2 eV[23]. When fitting the Cls peak of chlorinated carbon blacks, the best fitting was reached when the differ- ence between the components attributed to C-Cl and C-Cl, was equal to 1.2 * 0.2 eV.

3. RESULTS AND DISCUSSION

Figure 2 shows high-resolution fitting of chlorine (C12p) and carbon (Cls) spectra of the carbon black MT N 990, as well as digital difference spectra between Cls peaks of the initial and the chlorinated samples. The binding energies are referred to the 1s level of contamination carbon whose binding energy, as de- termined by the calibration of our spectrometer, is 285.2 f 0.1 eV.

3.1 Cl2p peak This peak is complex, since it is fitted by five com-

ponents (more exactly, doublets). Component (l), having a binding energy of 200.0 + 0.2 eV, is attrib- uted without doubt to Cl atoms covalently bonded to sp2 and sp3 carbons. One notices that the binding en- ergy is somewhat lower than for Cl bonded covalently to sp2 carbon in chlorinated polymers (i.e., 200.45 -t 0.1 eV), and to the value we recorded with our refer- ence chlorinated compounds (i.e., 200.6 + 0.2 eV). This difference is either due to the different position of the Fermi level or to the influence of delocalized K electrons[24], which cause slightly higher electron den- sity around the Cl nucleus of the Cl bonded to carbon black.

Component (2), shifted 1.10 f 0.2 eV from (1) in the direction of increasing binding energies, has a binding energy of 201.10 i 0.2 eV. Nefedov[ZS] aIso observed a difference of the same order of magnitude between the binding energies (C12p 3/2) of chlorine in monochlorinated compounds compared to Ccl4 (solid). In our case, since the chlorination temperature was 450°C it seems hard to believe that Ccl, mol- ecules remained unreacted on the carbon surface. The relative intensity of compound (2) is not appreciably higher in the spectra of the high-surface-area chlori- nated XE-2 black. Hence, this form of Cl is certainly bonded chemically, probably in the form of C-Cl, groups linked to peripheral aromatic carbon atoms of the black (Ccl, substitution of H atoms). However, there is no evidence of this possibility in the literature, nor did we find a model compound to directly verify our hypothesis. The “structures” of the fitted Br3p and Br3d peaks of the brominated carbon blacks[22] and

of the C12p peak of the chlorinated samples show close analogies. Besides the main component correspond- ing in one case to Br atoms and in the other to Cl at- oms, covalently bonded to carbon atoms of the carbon black, one distinguishes in each case components of weak binding energies corresponding to ionic Br and Cl species and a component of significantly higher en- ergy (3.5 eV and 4.0 eV away from the main compo- nent), which we attributed to occluded Br, and Cl* molecules. However, component (2), located between the main component (1) and the component of high binding energy of the C12p peak, and which is always more intense than the latter three components, is not differentiated on the Br3p and Br3d peaks. This dif- ference is explainable considering that in the chlori- nation experiment the reactive gas mixture contains, in addition to CIZ, noticeable amounts of Ccl,, gen- erating Ccl, radicals at 45O”C, whereas in the bromi- nation experiment it contains no polybrominated carbon groups.

Component (3), shifted 4.0 + 0.2 eV from (1) in the direction of increasing binding energies, corresponds to a binding energy of 203.9 f 0.2 eV. This position is not the one of oxychlorinated groups, since the binding energies of 2p 3/2 electrons in ClO; and ClO, are much higher. By analogy with the compo- nent (4) of Br3d and Br3p, we may attribute it to oc- cluded Cl,.

Component (4), shifted 2.3 + 0.2 eV from (1) in the direction of decreasing binding energies, is attributed to Cl- (chlorides, occluded HCl).

Component (5), shifted 3.3 * 0.2 eV from (1) in the direction of decreasing binding energies, is beyond the known energies scale of Cl- of chlorides. It may corre- spond to surfaces charge transfer complexes (C--t Cl,).

Components SUl and SU2 are possibly related to shake up (a + K*) satellites or plasmons.

3.2 Cls peak Six components are required to fit the Cls peak. Component (l), having a binding energy of 284.3 k

0.1 eV, corresponds essentially to non-functionalized sp2 and sp3 carbon atoms and, to a small extent, to nonfunctionalized sp2 and sp3 carbons in a 0 position of a fixed Cl or to nonfunctionalized aliphatic carbon atoms of contaminants (either on the carbon surface or, more probably, on the sample holder). The bind- ing energy corresponding to the envelope summit of the Cls peak, the width at mid-height, and the asym- metry of component (1) are the same for the initial and the chlorinated samples. Like bromination, chlorina- tion does not significantly affect the polyaromatic structural units and the delocalized 7r electron system (there is no appreciable aliphatization).

Component (2), shifted 1.2 -t 0.2 eV from (1) in the direction of increasing binding energies, is essentially attributed to sp2 carbon atoms linked to one chlorine atom (chlorine substituted sp2 carbon atoms at the pe- riphery of the polyaromatic structures).

Component (3), shifted 1.9 + 0.2 eV from (I), cor- responds to sp3 carbon atoms carrying a Cl atom (Cl

Page 5: XPS study of the halogenation of carbon black—Part 2. Chlorination

XPS study of the halogenation of carbon black-Part 2

Fig. 2. XPS spectra of chlorinated MT N 990. (a) fitted C12p peak, (b) fitted Cls peak, (c) numerical dif- ference between initial and chlorinated samples.

fixed by a free radical process on two nonconjugated Component (4), shifted 3.2 * 0.2 eV from (l), is sp2 carbon atoms) and, to a lesser extent, to sp2 and attributed to chlorinated carbon (Csp3 carrying 2 Cl sp3 carbons linked to a single 0 atom (C-OH, atoms or, eventually, aliphatic C with one Cl atom) ether . . .). and slightly to oxygenated groups (carbonyl).

Page 6: XPS study of the halogenation of carbon black—Part 2. Chlorination

68 E. PAPIRER et al.

Table 1. Elemental comoosition and composition detected by XPS (atomic ratios) of chlorinated carbon blacks

Samole Cl (%) [CI]/[C] from [CI]/[C] from 0 (Q) [O]/[C] from [O]/[C] from

weight Chem. XPS T” (070) weight Chem. XPS

Printex XE-2 1.31 0.0274 0.0248 3.7 1.49 0.0123 0.0058 Corax N 115 5.42 0.0196 0.0200 20.0 0.85 0.0068 0.0061 Corax N 326 6.92 0.0253 0.0300 41.0 0.67 0.0054 0.0040 Corax N 712 7.55 0.0278 0.0385 135.0 0.44 0.0036 0.0066 MT N 990 4.15 0.0147 0.0765 271.0 0.21 0.0017 0.0108

“Surface coverage by Cl

Component (S), shifted 4.1 + 0.2 eV from (l), is attributed to dichlorinated carbon atoms (aliphatic carbons with two Cl atoms, sp3 carbons, dichlorinated, linked to an already monochlorinated carbon atom) and to oxygenated groups (carboxyl).

Component (6), 5.5 + 0.2 eV away from (l), is more apparent on chlorinated than on the initial car- bon black samples. In the latter case, it is possibly due to shake-up satellites (r--t a*) and, eventually, to CO (adsorbed, occluded). On chlorinated samples, it cor- responds essentially to -Ccl, types of groups (sub- stitution of edge sp2 carbon atoms).

Component (PL), shifted 7.0 +- 0.3 eV and 8.3 + 0.3 eV, respectively, from (l), originates from plas- mons. Furthermore, it is important to underline that the tailing of the peak is only partly taken into account in the fitting procedure. Moreover, carbon-oxygen groups contribute also to components (3) to (5). Hence, the relative intensities of components (2) to (6) will only allow a semi-quantitative evaluation of the various types of carbon-chlorine bonds.

3.3 Numerical spectra substraction Figure 2c shows the digital difference spectrum ob-

tained by numerically substracting the Cls peak of the initial MT N 990 carbon black from the correspond- ing peak of the chlorinated sample. The spectrum ob- tained is then further analyzed and fitted using the individual components of the Cls peak (by fixing the positions and widths and mid-height and by adjusting the intensities). The following observations are made: The component (l’), with a binding energy of 284.8 + 0.2 eV, corresponds to non-functionalized sp2 and sp3 carbon atoms in p position to a Cl atom and to ali- phatic carbons of hydrocarbon contaminants or struc- tural aliphatic units of the black (with proportions differing slightly between initial and chlorinated black). The amount of C participating in carbon-oxygen groups being noticeably less important (on the surface or in the subsuperficial zone) in the chlorinated sample than in the initial sample, the components (2’) to (6’) will originate essentially from chlorinated carbons. Com- ponent (2’), having a binding energy of 285.3 ? 0.2 eV, corresponds to sp2 carbon atoms at the edge of the cy- cles where hydrogen has been substituted by chlorine. Component (3’), with a binding energy of 286.1 + 0.2 eV, is related to sp3 carbons carrying a Cl atom (Cl

fixed by free radical addition on nonsaturated carbons belong to (3’)).

Component (4’), having a binding energy of 287.35 -+ 0.2 eV, is attributed in part to sp3 carbons carrying two atoms of Cl and, in part, to monochlo- rinated aliphatic carbons.

Component (5’), with a binding energy of 288.3 +- 0.3 eV, and presenting a weak intensity, may corre- spond to aliphatic carbons carrying two Cl atoms.

Component (6’), with binding energies around 289.65 f 0.2 eV, is attributed to -CCls groups.

Finally, (PL) corresponds to plasmons (slightly more pronounced on chlorinated than on initial black).

The Cl atoms, bonded to C, which contribute to components (2’) to (5’) of the digital difference spec- trum, all contribute also to component (1) of the Cl2p peak. If one assumes that components (2’) to (5’) cor- respond solely to chlorinated carbons, the area of the component (1) of the peak C12p should be equal to:

Area C12p(l) = (58/20)

x [Area (2’) + Area (3’)

+ Area (4’) + 2 Area (5’)].

In the case of the chlorinated N 990, the calculated area is 1700 -t 20% cps eV (the error being larger for the areas of the digital difference spectrum than for the intensities of the separate spectra). The measured area is 1400 It_ 10% cps eV (i.e., a value in agreement with theoretical expectations).

If component (6’) is essentially due to --Ccl,, then the area of component (2) of the C12p peak, which was attributed to Cl from -Ccl,, should be of the order of: Area C12p (2) = (58/20) x Area (6’), that is, 370 f 40% cps eV. The actual value is 200 * 10% cps eV, a value which supports our evaluations.

Table 1 reports the contents of chlorine (weight%) in chlorinated carbon blacks (column I), the atomic [Cl]/[C] ratio determined either by elemental analy- sis (column 2), or from XPS spectra (column 3), the surface coverage r (column 4), the weight percentage of oxygen, measured by chemical analysis (column 5), and the atomic ratios [O]/[C] evaluated, respectively, by elemental analysis (column 6) or XPS (column 7).

Table 2 indicates the relative contents of the vari-

Page 7: XPS study of the halogenation of carbon black—Part 2. Chlorination

XPS study of the halogenation of carbon black-Part 2

Table 2. Repartition of C12p components in chlorinated carbon blacks

Sample C12P (1) C12P (2) C12P (3) C12P (4) Cl2P (5)

69

Printex XE-2 70.7% 16.6% 4.8% 5.4% 4.2% Corax N 115 73.9% 14.3% 4.1% 5.9% 1.8% Corax N 326 72.2% 16.3% 2.1% 6.4% 3.0% Corax N 772 75.7% 14.3% 1.4% 6.6% 2.0% MT N 990 80.0% 11.4% 1.6% 5.2% 1.8%

ous types of chlorine bonds discussed earlier. For an easier comparison, Fig. 3 shows the ratios [Cl]/[C] de- termined, respectively, from chemical analysis and XPS spectra of the different samples. Figure 4 shows the ratios ([O]/[C]),,,, and ([O]/[C]),,s of the chlo- rinated samples, and Fig. 5 shows the corresponding ratios, respectively, in the initial and chlorinated samples.

Figure 6 displays fitted digital difference, spectra (numerical difference between the Cls peaks of initial and chlorinated carbon blacks). Even though the areas are not known with precision (k30 to 40%), it is ob- vious that the thermal black behaves differently than the other blacks upon chlorination, since the compo- nent (2’) attributed to chlorinated sp2 carbons (sub-

stitution of H by Cl) is significantly more important, compared to (3’) and (4’), which correspond to chlo- rinated sp3 carbon atoms. The high-resolution spec- tra of chlorine (C12p peak) from the different carbon

blacks are compared in Fig. 7. Figure 8 shows the valence region photoelectron

spectra of T.N.H.C.P. and of the initial and chlori- nated XE-2 and MT N 990 carbon black samples (the expanded spectra, for the energy window 1240 5 Ekin I 1265 eV are also shown for the MT N 990

samples). It is seen that the binding energies of the main component of the C13p peak in the spectrum of T.N.H.C.P. (Eb C13p = 5.4 f 0.1 eV) and the one at the summit of the band C13p on the spectrum of the chlorinated carbon black (&, C13p = 5.1 * 0.1 eV) are similar, which indicates that chlorine is essentially chemically linked. One may also note that the Fermi

0,oa

0,07

0,06

0,05

a,04

0,03

0,02

0.01

0

n [Cly[C] them. q [CI)/[C] xps

XE-2 N 115 N 326 N 772 MT

Fig. 3. Comparison of the atomic ratios [CI]/[C] determined either by chemical or XPS analysis.

edge is clearly more pronounced in the spectra of chlo- rinated carbons. Clearly, the density of states in the neighbourhood of the Fermi level increased after chlori- nation. A similar observation was made by Sch16g1[6], who studied chlorinated graphites (intercalated with FeCl,) in comparison with the initial HPOG.

4. CONCLUSIONS

Carbon blacks fix much more chlorine than bro- mine, a result that is in agreement with earlier find-

ings. For instance, XE-2, N 115, and N 326 fix twice as much or more chlorine than bromine (in number of halogens per 100 atoms of C, as determined from chemical analysis). The amount of fixed chlorine is 4

0,03-

0,025-

0.02-

0,015-

O.Ol-

7

XE-2 N 115 N 326 N 772 MT

Fig. 4. Comparison of the atomic ratios [O]/[C] determined by XPS on initial and chlorinated carbon blacks.

0,025

0,02

0,015

q [O)I[C] them. Chlor.

q [OY[C] xps lnltkl

XE-2 N115 N 326 N 772 MT

Fig. 5. Comparison of the atomic ratios [O]/[C] on initial and chlorinated carbon blacks determined either by chemi-

cal analysis or by XPS.

CAR 33:1-D

Page 8: XPS study of the halogenation of carbon black—Part 2. Chlorination

70 E. PAPIRER et al.

[CPSI 4

I 3

960 964

Kin. energy h ] 972

Fig. 6. Digital difference spectra (chlorinated-initial) of car- bon black samples.

times higher in N 772 and 8 times higher in MT N 990. ysis or XPS, it appears that these ratios are practically The major part (84-91%) of chlorine is covalently of the same order for XE-2 and N 115, and differ only bonded to sp2 and sp3 carbon atoms of the black. The slightly for N 326. The difference becomes more pro- weight percentage fixed shows no relation to the sur- nounced in N 772, and is maximal for MT N 990. face area of the blacks. When comparing elemental However, with the two last blacks, this difference is

N772

4

L c

4

i

5 5

(

L

L-

\,. ioss

Kin. energy [ eV ]

Fig. 7. High-resolution spectra of chlorine from the differ- ent carbon blacks.

compositions determined either from chemical anal- significantly lower than in the corresponding bromi-

Page 9: XPS study of the halogenation of carbon black—Part 2. Chlorination

XPS study of the halogenation of carbon black-Part 2

Cl 3p

71

(: 2s Cl 3s I

.l

1225 1230 126 1240 1245 1250

Fig. 8. Comparison of XPS spectra of reference product and initial and chlorinated carbon black samples.

Page 10: XPS study of the halogenation of carbon black—Part 2. Chlorination

72 E. PAPIRER et al.

nated samples. For instance, in MT N 990, the ratio (ICl]/[C]),,, is 5 times higher than the ratio ([Cl]/

[clkhem s whereas the ratio ([Br]/[C])xPs was 28 times higher than the ratio ([Br]/[C])ch,,,. The chlorination reaction proceeds inside the carbon black particles at distances larger that those observed during bromina- tion. By assuming that Cl is uniformly distributed in a shell of thickness 6 and by application of the relation 6 = d/2[1 - (1 - cu)“3], given in ref. [22], where d represents the mean diameter of the particles and D( = ([Cl]/[C]),,,,/([Cl]/[Cl),,,, one calculates 6 value equalto9&2nmforN326and6=11 It_2nmfor the thermal black.

For the thermal black, chlorination occurs princi- pally through a hydrogen substitution process. More- over, the 0 content of the carbon black is significantly decreased, suggesting a chemical reduction of the ox- ygenated surface groups, which are possibly replaced by chlorine atoms.

In conclusion, it is shown that a combination of chemical and spectroscopic methods leads to a more precise and quantitative approach to the complex chlo- rination process of carbon blacks,

6. 7.

8.

9.

10. 11.

12.

13.

14. 15.

16.

17. 18.

19.

20.

REFERENCES 21. 22.

J. 0. Hooley, Can. J. Chem. 35, 1414 (1957). R. SchlBgi, Surface Sci. 189/190, 861 (1987). .I. B. Donnet and R. C. Bansal, In Carbon b/ark science and technology (Edited by J. B. Donnet, R. C. Bansal, and M. J. Wang), pp. 183-186. Marcel Dekker, New York (1993). H. P. Boehm, U. Hofmann, and A. Clauss, Proc. Third Carbon Conf., Buffalo, NY. Pergamon Press, NY (1957). B. R. Puri, S. L. Malhotra, and R. C. Bansal, J. Ind. Chem. Sot. 40, 179 (1963). B. R. Puri and R. C. Bansal, Carbon 5, 189 (1967). B. R. Puri and R. C. Bansal, Ind. J. Chem. Sot. 5, 381 (1967). B. R. Puri, 0. P. Mahajan, and D. D. Singh, J. Ind. Chem. Sot. 38, 943 (1961). B. R. Puri, N. K. Sandle, and 0. P. Mahajan, J. Chem. Sot. 4880 (1963). B. R. Puri and R. C. Bansal, Carbon 3, 533 (1966). B. R. Puri, D. L. Gandhi, and 0. P. Mahajan, Carbon 15, 173 (1977). B. R. Puri, D. D. Singh, and V. M. Arora, J. Ind. Chem. Sot., 488 (1978). B. R. Puri and K. C. Seghal, Ind. J. Chem. 5,379 (1967). D. L. Gandhi, S. K. Sharma, A. Kumar, and B. R. Puri, Ind. J. Chem. Sot. 13, 1713 (1975). B. R. Puri, 0. P. Mahajan, and D. L. Gandhi, Ind. J. Chem. 10, 848 (1972). L. H. Reyerson and A. W. Wishart, J. Phys. Chrm. 42, 679 (1938). H. Tobias and A. Soffer, Carbon 23, 281 (1985). E. Papirer, R. Lacroix, J. B. Donnet, G. NansC, and Ph. Fioux, Carbon 32, 1341 (1994). 0. Beamson and D. Briggs, In High resolution XPS of organic polymers. The Scienta ESCA 300 dala base. Wiley & Sons, New York (1992). A. Proctor and P. M. A. Sherwood, J. Electron Spec- Irosc. 27, 42 (1982).

G. Hennig, J. Chem. Phys. 20, 1443 (1952). 23. R. Juza, P. Schmidt, A. Schmeckenberger, and P. Jonck, Naturwissenschafren 42, 124 (1955). R. Juza and H. Seidl, 2. Anorg. Allgem. Chem. 317, 73 24. (1962). R. Juza, P. Jenck, and A. Schmeckenberger, 2. Anorg. 25. V. I. Nefedov and M. A. Ponaikoshits, Mater. Res. BUN.

Chem. 292, 34 (1957). 7, 1543 (1972).