(hrbon Vol 27. No. 6. pp. YSI-YhS. Pruned in Great Britain
IXIOX-6223/X9 53.Ml + .O(I P 1989 Pergamon Press plc
19x9
LETTERS TO THE EDITOR
Reaction of an Fe-graphite intercalation
compound with FeCl3
(Received 25 May 1988; accepted in revised form 24 February 1989)
Key Words - Graphite intercalation compounds, iron chloride-graphite, reduction of FeCl3-graphite, iron graphite
Iron-GICs catalysts for hydrogenation reactions are obtained by reduction of FeC13-GICs. Such catalysts with a near-complete filling of the interlayer spaces have not been available so far because the highly reactive reducing agents lead to partial (or complete) loss of iron chloride from FeClj-GIC. Pritzlaff and Stahl (1) studied the reaction of iron (III) chloride GIC with Fe(CG)s in the presence of CO. According to this method an FeC13-GIC can be reduced to an FeClz-GIC at relatively low temperatures by insertion of additional Fe atoms into vacant octahedral cation sites of the intercalated FeC13 layers: 2 FeCI,-GIC
+ Fe(CO)s + 3 FeC$-GIC
+ 5 CG
The Fe(CO)S as reagent cannot penetrate directly into the unfilled interlayer spaces of bulk graphite. We have proposed an alternative to Pritzlaffs synthesis of FeClz-GIC means of reaction of FeCl3 vapour with solid Fe-GIC. In our method, the vapour of Fe03 can additionally and directly intercalate the unfilled spaces between the carbon layers of graphite. The product of synthesis is the starting material for reduction, and it serves as a stable precursor of a catalyst with a high concentration of intercalated iron. The FeC13-GIC samples were preparedaccording to the commonly used method of Rbdorff and Schuli (2). Powdered natural Sri Lanka aranhite with a narticle 1 ~~ size of l-20 microns in thickness and 30-100 microns in width was used in this study (Fig. 1). The mixture of powdered graphite and pure anhydrous ferric chloride (Riedel) was heated at a temoerature of 300°C for 20 h. The products of intercalation were freed from excess metal halide by washing with an aqueous solution of HCl (l:l), filtering, washing with distilled water, and drying overnight at 110°C. The reduced samples were obtained by polythermal reduction of the GIC with a mixture of nitrogen and hydrogen (1:3) under atmospheric pressure at temperatures from 15O’C to 3CVC with a heating step of 2YC per 24 h, and from 300°C to 625’C with a heating step of 50°C per 24 h followed by further reduction at 625°C for 5 days with a space velocity of the gases of ca. 1000 h-l. The a-iron which partially appeared during the reduction process was removed by an aqueous solution of HCl (1: 1). The remaining mixture of GICs was used for the re-intercalation of FeC13. The history of the treatment of the samples is given in Table 1. The “training” of GIC can be Y
1
Fig. 1. Photomicrograph of natural Sri Lanka graphite. described in the following steps: 1) Intercalation, 2) Reduction, 3) Removal of u-Fe 4) Intercalation, 5) Reduction, 6) Removal of u-Fe 7) Intercalation, 8) Reduction, 9) Removal of a-Fe. The products of preparation have been characterized by X-ray fluorescence spectroscopy and XRD data. The measurements were made by means of a UniversalRdntgen-Diffractometer HZG-4 (GDR) with MoKa radiation and an X-ray fluorescence spectrometer VRA30 (GDR) using the Kat line of chlorine and the KPl line of iron. Novikov et al. (3) found that no Fe-GIC is formed on heating of FeClJ-GIC in a stream of Hz at temperatures of 375-400°C for 4 - 12 h. MZSssbauer spectroscopy studies on the resulting material (3,4) confirmed formation of FeClz-graphite. In the present work we have applied higher temperatures and a longer time of reduction. 951
952
Letters to the Editor Table 1
Step
Preparation
X-ray and elemental analyses of the studied samples.
HCl treat
%Fe
intercalationof graphite
+
17.6
reduction of sample No. 1
-
23.0
12.1
64.9
13.12
0.82
G a-Fe Fe-GIC St. 1 (585 pm) FeCI2-GIC(1300 pm-weak)
washing of sample 2 with HCI solution
+
19.1
15.2
65.7
16.0
1.25
G Fe-CIC FeQ-GIC St.2 (1300 pm-weak)
intercalationof sample No. 3
+
27.6
24.1
48.3
8.13
1.37
G FeC12-GICSt. 1 (960 pm) FeClJ-GIC st. 1 (936 pm)
reduction of
-
30.3
17.1
52.6
8.07
0.88
G a-Fe Fe-GIC St. 1 (582 pm)
99Opm
+
28.1
18.2
53.7
8.89
1.02
G Fe-GIC St. 1 (582 pm)
99Opm
%Cl
%tz
C/Fe
idetuitied 32.2
50.2
13.26
2 8X -.__
001 series not identitied
c-
FeCl&IC St. 2 (1275 pm) + St. 1 (935 pm)
sample No. 4
waallingof sample5 with
Cl/Fe
HCl solution intemalationof sample No. 5
-
reduction of sample No. 7
-
31.1
16.6
52.3
7.82
0.84
G Fe-!IC St. 1 (589 pm)
washing of sample 8 with HCI solution
+
29.5
17.7
52.8
8.33
0.95
G Fe-GIC St. 1 (589 pm)
* balance to 100%
28.6
23
48.4
7.87
1.27
G FeClr-GIC St. 1 (960 pm) Fe-GIC st. 1 (580 pm)
915 pm
** G-graphite : The main constitutena are shown in hold type
The diffractograms of samples 1,2,3,4,5,7 and 8, are presented in Fig. 2. The results of the elemental analyses and X-ray diffraction data before and after intercalation, reduction, removal of a-Fe, and washing with hydrochloric acid are presented in Table 1. The main constituents are shown in bold type. They were determined by measuring the intensities of the X-ray reflections. In the samples obtained by intercalation of FeCl3 vapour one can observe the formation of a periodic structure with a distance between the graphene layers from 935 pm (FeCl3GIC) to 960 pm (FeCl&IC) as main constituents, and at higher iron concentrations with ca. 990 pm as a secondary constituent and 915 pm as the main constituent. The growth of the distance between the graphene layers as: 935 pm + 960 pm + 990 pm is probably caused by an increase of the Cl-Fe distance. The decrease of the distance to 915 pm is presumably due to a distortion of the octahedral chlorine coordination of the iron centres, and/or formation of an non-stoichiometric iron-chlorine compound between the graphene layers. In the reduced samples one can notice the formation of an iron-GIC with an interlayer distance of 582-589 pm as main constituent. The elemental analysis of these samples yield a ratio Cl:Fe = 0.8. This means that some chlorine compounds remained in the samples. But yet the characteristic XRD reflections for phases of both FeClz-graphite and FeCl2 were not observed. For
FeCl2 the (104) reflection (254 pm) is more intensive (100%) as compated to the (003) reflection (590 pm) (63 %; ASTM card l-1106). The chlorine species are probably bound in the intercalation compound in a disordered form. The molar ratio C:Fe = 8 in samples No 4-9 is approximately constant. The range of C:Fe values is in reasonable accordance with the intensities of the graphite (002) XRD reflections (Fig. 2). The diffractograms No. 5 and 7 (Fig. 2) indicate that in the course of intercalation part of the FeCl3 vapour reacts with a-Fe located on the graphene surface to form the non-intercalated FeCl2 which can be removed by water. We have discovered the possibility of reacting solid Fe-GIC with FeCl3 vapour. Depending on the C:Fe ratio in the starting material this reaction of intercalation of FeCl3 in Fe-GIC as main constituent can be described as follows:
Fe - GIC + FeCl3 V
p
FeClx - GIC (x4) LFe
- GIC
Letters
YS3
to the Editor
No7
No 4
No 3
The intercalation of FeC13 did not cause deintercalation of iron from the Fe-GIC, but additionally the vapour of FeC13 was penetrating directly in the unfilled interspaces between the graphene layers, in addition. At lower iron contents the formation of a FeCljGIC prevailed on intercalation. At higher iron contents in the starting mixture the process of oxidation of the iron-GIC to FeClz-GIC or non-stoichiometric FeCl,GIG dominated and the phase of free graphite disappeared completely. During the reduction of FeClj-GIC samples by 3H2 + N2, a mixture of new phases appeared: i) iron-GIC (d = 578 - 589 pm). The additional appearance of series of integral (OOQreflections requires explanation. ii) a-iron (d = 202 pm) on the surface of the graphite particles. iii) free graphite (d = 335 pm) with XRD intensities depending on the concentration of iron. During intercalation, the a-iron is removed from the surface of the carbon particles: a-Fe/graphite + FeC13 + FeCldgraphite. The FeClzformed on the surface is washed off with water in the course of preparation. After reduction, the a-iron can be removed by diluted hydrochloric acid. In spite of the fact that chlorine was detected in reduced and HCl-washed material the phases of both FeC12 and FeClZ-graphite were not observed. Therefore the reduced samples contained unidentified chlorine compounds. The presented procedures indicate the possibility of preparing a mixture of stages of iron-GIC with the 1st stage dominant. The obtained material forms a stable precursor of a catalyst with a high iron content. Obviously, in unwashed samples admixtures of small amounts of FeC12 are present.
No2
I
I
Institute of Chemical Tech. Technical University of Szczecin ul. Pulaskiego 10 70-322 Szczecin, POLAND
,
I
.
I
I
REFERENCES
No1
,
30'
20”
10”
20
1: 3.
Fig. 2. Diffratograms of samples of No 1, 2, 3, 4, 5, 7 and 8. MoKa radiation, “d” values in picometres
A. W. MORAWSKI K. KAtUCKI
4.
B. Pritzlaff and H. Stahl, Carbon, 15,399 (1977). W. Rudorff and H. Schulz. Z. Anorg. Allg. Chem., 245, 121 (1940). Ju. N. Novikov, M.E. Volpin et al., Zhurnal Strukturalnoi Chimi, 11, 1039 (1970). J.G. Hooley, M.W. Bartlett, B.V. Liengme and J.R. Sams, Carbon, 6,681 (1968).