New biintercalation compounds in a potassiumcopper chloride-graphite system

Carbon, Printed

Vol. 31, No. 8, pp. in Great Britam.

1325-1331.

1993 CopyrIght

NEW BIINTERCALATION COMPOUNDS COPPER CHLORIDE-GRAPHITE

WX6223/93 C 1993 Pergamon

$6.00 t .W Pres\ Ltd.

IN A POTASSIUMSYSTEM

T. Yu. SOKOLOVA,~ V. A. NALIMOVA,’ 0. V. FATEEV,’ V. V. AVDEEV,’ K. N. SEMENENKO,’ and D. GUI~RARD~ ‘Department of Chemistry, Moscow State University, Leninskie Gory, Moscow, 119899, Russia ?Topchiev Institute of Petrochemical Synthesis of Russian Academy of Sciences, Leninsky pr., 29, Moscow, Russia 3Laboratoire de Chimie du Solide MinCral, UniversitC Nancy 1, CNRS no. 158, BP 239, 54506 Vandoeuvre-Its-Nancy, Cedex France (Received

30 April 1993; ucceppted in revised form 9 June

1993)

Abstract-Potassium interaction with stage 1 and stage 2 CuCl,-graphite intercalation compounds (GICs) at T = 200°C was investigated. Two new phases with I, = 14.70 8, and 12.55 A were fixed. The analysis of the intensities of 001 reflections has shown that these two phases seem to be biintercalation compounds with the following sequences of the layers: K\CuCl\K\CuCI, respectively. Key Words-Graphite,

potassium, biintercalation.

copper chloride. structure

1. INTRODUCTION

Biintercalation compounds containing a sequence of “donor” and “acceptor’‘-type layers in graphite matrix should be very interesting from the point of view of their physical and chemical properties. In 1985, Suzuki, Chow, and Zabel realized the intercalation of potassium into the stage 2 CoCl,-GIC[l]. X-ray diffraction studies have shown the samples obtained being a mixture of two phases: initial CoCl,-GIC and new structure with I, = 14.65 A. Analysis of 001 reflection intensities has shown that the new phase is a biintercalation compound with the following sequence of layers: CoC12\K\CoC12\K (\-graphite layer). It was noted that potassium atom concentration in the layer is KC,* or KCl,. Further interaction of potassium with this system has led to the formation of the mixture of stage 1 CoCl,-GIC and stage 1 K-GIC. This reaction was explained with the help of the Daumas-Herold model. A number of heterostructures were obtained by Furdin et a1.[21 by interaction of stage 2 MCI, (M = Co, Cd) with alkali metals; their repeat distances are listed in Table 1. These repeat distances are rather close to the values calculated as a sum of the repeat distances of stage 1 CdCl,-GIC and stage 1 alkali metal-GIC. Some differences between calculated and experimental values were explained by partial reduction of CdClz inside host graphite; the thickness of species (CdCl, and Cd”) is less than that of CdCl, intercalated between graphene layers. And the total reduction of transition metal and the formation of alkali metal chloride has happened under harder conditions. Analogous results were obtained in systems: NiClz - GIC - M (M = Li, Na, K)[3], FeCl,GIC - K[4,5], CoCl,-GIC - K[4]. The authors of refs. [6] and [7] fixed transition metal chloride reduction and formation of transition metal layers CAft31:8-E

between graphite sheets under the same conditions. Not only stage 2 acceptor type GICs may be intercalated by alkali metal. The reaction between stage 4 CdCl,-GIC and sodium was carried out by authors of ref. [2]. It was found out that alkali metal occupies all vacant graphite galleries, whereas the interaction of graphite with sodium in the same way leads to high-stage (6 or 8) GICs. Murakami et a[.[81 have investigated the interaction of stage 3 and stage 4 MoCI,-GICs with potassium. It was shown that at first potassium entered into one vacant graphite gallery next to MoCl, layers. Then potassium occupies another gallery. Intercalation of acceptor-type GICs by alkali metal is the only way to synthesize donor-acceptor biintercalation GICs. The attempts to realize a reaction between donor type GICs and transition metal chloride resulted in the reduction of chloride without intercalation[4,6,9,10]. CuCl,-GICs were not investigated in this type of reaction, and we were mainly interested in the details of the reaction of potassium with CuCl? in a quasi two-dimensional state. 2. SYNTHESIS

For the CuCl,-GIC synthesis we used disks (D = 6 mm, thickness 0.3-0.4 mm) of highly oriented pyrolytic graphite (HOPG). CuCl,-GICs were synthesized in the two-bulb quartz ampoules under chlorine atmosphere. The GIC stage number was determined by X-ray diffraction (Cu K, radiation), and composition was determined by mass increase. For stage 1 CuQ-GIC synthesis both graphite and CuCI, were placed into the same section (T610”C). The time of synthesis was two weeks. CuCI, excess was removed to the cooler section (T 600°C). Composition varied from C,,9CuCl, up to C5,1CuCI?. I, = 9.40 A; diffractogram is shown in Fig. la.

I325

T. Yu. SOKOLOVA etal.

1326

Table 1. I, values for M - CdClz - GICs Z, calculated,

Metal Na K Rb cs

8,

Z, experimental,

14.08 14.90 15.20 15.51

A

13.80 14.90 15.25 15.10

Stage 2 C&l,-GIC synthesis was carried out under gaseous phase conditions at the same temperatures (Z.. = 610°C Tchloride= 6OO”C),synthesis time being four weeks. The composition varied from C,,,CuC12 up to C,,,$uC12, Z, = 12.75 A (Fig. lb). The interaction of CuCl,-GICs with potassium was carried out into the two-bulb Pyrex tube under vacuum at 200°C with temperature gradient not more than 10°C. All the samples were transferred in the glove box, in dry inert (argon) atmosphere.

3. RESULTS AND DISCUSSION

Different products of the reaction of potassium with stage 1 and stage 2 CuC12-GICs at 200°C are listed in Table 2. The reaction was stopped at a definite time for X-ray diffraction study of the sample. The sample was examined from the surface (Table 2, outside), and then split and examined inside (Table 2, inside). It is easy to see from Table 2 and Fig. 2 that

1

50

I

45

potassium intercalates into stage 2 CuC&-GIC, forming the compound with Z, = 14.70 A. We have fixed formation of this phase on the surface of the sample after one hour of the reaction. After five hours of the reaction we observed 70% of the phase with Z, = 14.70 A inside of the sample (see Table 2). This value of Z, is a little less than the sum: 9.40 A (Z, of stage 1 CuCI,-GIC) + 5.35 A (Z, of stage 1 K-GIC) = 14.75 A. So our supposition was that this phase corresponded to the biintercalation compound CuCl,\K\CuCl,\K . . . O-graphite layers). To establish the positions of atomic planes along the z-axis we have done quantitative analysis of the intensities of 001 reflections (Table 3). Structural factors obtained and used for the Fourier transform give a profile of electron density p(z), shown in Fig. 3. It is in good agreement with the profile calculated on the basis of the centrosymmetric model, where zc, = 0; zc, = 20.097; zc = ko.3177; z, = kO.5 for the theoretical composition C~.~CUC~~K~,~ (Fig. 3). Residual factors (calculated for both the intensities and structural factors) are low enough. The structure of the CuC12 layer is close to those proposed by A. Perignon etaZ.[ 1l] for the stage 2 CuCl,GIC (Fig. 4). Figure 4 shows also an electronic profile for the stage 1 CuCl,-GIC. The stackings of Cu and Cl layers are similar in both compounds. So the structure of the CuCl, layer in the biintercalation compound with potassium seems to be the same as in pure stage 1 and stage 2 CuCI,-GICs. The stoichiometry of the potassium layer was close to KC*.

L

10

3.5

30

Fig. 1. CuC12-GIC diffractograms:

25

t

20

(a) stage 1; (b) stage 2.

15

10

5

New biintercalation Table 2. Interaction

1327

compounds

of potassium with stage nCuClz-GlCs Quantity outside (%)

Quantity inside (%)

4 n

Compound

2

Cl, lCUCl2 phase- I phase-2 KCI, Cu KC,

Time, hours

A

1

3

5

16

48

1

3

5

16

48

12.75 14.70 12.55

100 -

60 40 tr. -

30 70 tr. tr. -

20 40 40 tr. tr.

55 tr. 45

90 IO -

tr. IO 90 tr. tr.

10 90 tr. tr.

80 tr. 20

70 tr. 30

5.35

Time, hours

cs,CuQ

I

phase-2 KCI, Cu KC,

9.40 12.55 5.35

1

3

5

18

51

1

3

5

18

51

100 -

100 -

10 90 tr. -

100 tr. -

85 tr. 15

80 20 tr. -

80 20 tr. -

5 95 tr. tr.

90 tr. 10

85 tr. 15

a

b 40

3.5

30

25

YB

15

10

5

Fig. 2. Stage 2 CuC&-GIC diffractogram after five-hour interaction with potassium at 200°C (inside): * = phase with I, = 14.70 A; (001)~stage 2 Cm&-GIC.

T. Yu. SOKOLOVA et

1328

Table 3. Calculated and experimental intensities of (001) reflections of biintercalation compounds in CuCl,-Kgraphite system Heterostructure 001

0 (talc.)

001 002 003 004 005 006 007 008 009 0010 0011 0012 0013 0014

3.004 6.015 9.044 12.099 15.188 18.324 21.518 24.783 28.137 31.600 35.196 38.960 42.936 47.187

with I, = 14.70 A

dool (talc.) 14.700 7.350 4.900 3.675 2.940 2.450 2.100 1.837 1.633 1.470 1.336 1.225 1.131 1.050

I ev’

I cd2

*I

44.72 8.38 100.00 1.08 35.16 23.03 1.08 0.32 14.62 18.92 0.29 7.99 2.28 0.23

6.2 100.0 1.7 34.9 13.9 1.2 1.2 16.0 17.2 0.0 4.9 1.8 0.5

R = 9.5%. Heterostructure 001

0 (talc.)

001 002 003 004 005 006 007 008 009 0010 0011 0012

3.519 7.051 10.610 14.211 17.871 21.608 25.444 29.406 33.530 37.861 42.464 47.434

R = 9.8%. * Experimental

with I, = 12.55 A I exP

&,(calc.)

*

12.550 6.275 4.183 3.138 2.510 2.092 1.793 1.569 1.394 1.255 1.141 1.046

0.0 8.0 100.0 0.0 * 3.5 10.7 0.5 0.6 0.6 1.6

i

25.48 1.33 19.19 100.00 0.01 0.03 3.35 10.76 0.53 0.00 0.54 1.69

values substituted by calculated ones.

600 ;

-G g 400

I talc

al.

This biintercalation compound was not stable at further contact with potassium, and after three hours of the reaction we saw close to the surface of the sample traces of the products of reduction of CuCl,, and also KCs. At the same time we observe a new phase with Z, = 12.55 A (Fig. 5). The same phase is observed at the reaction of the stage 1 CuCl,-GIC with potassium after five hours. The content of this phase grows up to 100% after 18 hours of the reaction, then KCs formation starts. This phase probably also corresponds to the biintercalation compound, but CuCI, in the layer is partly reduced, so the sequence of the layers is as follows: K\CuCl\K\CuCl . . . Profile of the electronic density p(z) obtained by Fourier transform of the structural factors calculated from the experimental intensities corresponds to that calculated on the basis of the proposed mode1 (Fig. 6). The 006 reflection corresponds also to the main line (111) of metallic Cu (d = 2.088 A); this coincidence explains why we have used the theoretical value instead of the experimental one (10%) of 006 reflection. It should be noted that C-K-C stacking thickness in this compound is larger (5.9 A) than in KC, (5.34 A). The calculation has shown that this increase in thickness is probably due to the splitting of the potassium layer along the c-axis, and the concentration of potassium in the layer is higher than in KC,. Then, the layer containing the copper chloride is quite thin: about 6.7 A, which is close to the normal thickness of chlorine GIC[ 121 or compounds presenting single intercalated layers of some transition metal chlorides, like, for instance, AuC1,[131. Li interaction with CuCl,-GICs proceeds under much harder conditions: 350°C in liquid phase and always accompanied by CuCl, reduction. At the same time, there is formation of a phase with Z, = 11.35 A (Fig. 7). This one seems to be a biintercalation compound with a partly reduced copper chlo-

CU

C

4 .i? 200 ; : E () 1

_2()() j I0

- I0 d fA) Fig. 3. Electronic density profile for C,,&uCI,Ko6.

New biintercalation

1329

compounds

.-+-

cu 400

-

‘;i 300

-

exp. talc.

c, z

200

-

i ~looL 2

o-

-z & loo-

a

-6 600 -

5

400

-

2 3 & 200

-

0

4

C

2

cu

6

“d&J C

5 2

O-

‘;ii c-200 Ql 2

-

--4oo I b

-10

Fig. 4. Electronic

density

profile

for (a) stage

1

1

0

10

d 1 C( &X12,

&I

(b) stage 2 C,” zCuClz

Fig. 5. Stage 2 CuCl,-GIC diffractogram after five-hour interaction with potassium phase with I, = 14.70 A; @I)-phase with I, = 12.55 A.

at 200°C (outside):

1330

T. Yu. SOKOLOVAet al.

CaLc. exp.

-

-200

1 -7

I -5

1 -3

I -1

1 1

I 3

b 5

1 7 d (.i,

Fig. 6. Electronic

density profile for C16,,(CUCI),,,K~,~.

c t

so

I

4

I

4

45

40

35

30

Fig. 7. Stage 2 CuC12-GIC diffractogram

1

25

L

\

I

I

20

15

IQ

after interaction with lithium.

New biintercalation

ride layer, as in the case of the hiintercalation phase with potassium whose 1, = 12.55 A. These data show the possibility of Cu( 1) chloride layers formation and conservation in graphite matrix in the reaction with potassium and lithium.

compounds 2.

G. Furdin, L. Hachim, D. Gu&ard, and A. HCroid, C. R. Acud. St?. Paris 301, serie 1, n 9, 579 11985).

3. R. Erre, F. BCguin, D. GuCrard, and S. Flandrois. Proc. C~rhon ‘86, Baden-Baden, p. Slh. 4. R. Erre. A. Messaoudi, and F. BCguin. .S.v,7rh.Mrr.

23, 493 (1988). 5. M. J. Tricker,

4. CONCLUSION

6.

Two phases were obtained by interaction of potassium with stage 1 and stage 2 CuCl,-GICs. The first one (1, = 14.70 A) is a “normal” biintercalation compound with the following sequence of layers: K\CuC12\K\CuClz. The second one (Zi = 12.55 A) seems also to be a biintercalation compound, but with layers of partly reduced CuCI,. the sequence of the layers being K\CuCl\K\CuCI.

7. 8.

9.

10. II.

REFERENCES 1. M. Suzuki, P. C. Chow. and H. Zabel, Piys. Rw. R 32, 6800 (I 985).

1331

12. 13.

E. L. Evans. P. Cadman. and N. C. Davies. Curhnn 12, 499 (1974). K. Klouda. F. Lizy. and I. Spalova, N&OU N 4, 21 (1981). Yu. N’. Novikov and M. E. Vol‘pin, Z/I. C’.FU.Kl~i~r. Oh.sc~hrsrucrN 1, 69 (1987). Y. Murakami. T. Kishimoto. H. Suematcu. K. Nishitani. Y. Sasaki. and Y. Nishina. Svnrh. Mtzr. 34, X5 (19891. A. H&old, G. Furdin, L. Hachim. N. E. Nadi. and R. Vangt!iisti. Anw. tic Phys. cd. n 2. *;up. n”. 11, 3 (I986). M. Inagaki. H. Mine. and M. Sakai, ./. .Sc,i. Wtrrca,.. sci. Jrq?. 37, 51 (1988). A. PCrignon, P. Pernot, M. Lelaurain. and K. Vang& listi. Cohort 27, 295 (1989). G. Furdin, M. Lelaurain, E. McRae, J. F. Mareche, and A. H&old, fwhon 17, 329 (1979). R. Vang&ti and A. H&old, C‘. K. AMY. SC Ptrr-i.c 276C, I109 (1973).