A new pathway to Mg2Ni and an MgNi carbide

Journal

of the Less-Common

A NEW PATHWAY

Metals,

163

- 172

163

TO Mg,Ni AND AN Mg-Ni CARBIDE*

B. BOGDANOVIC,

K.-H. CLAUS,

Max-Planck-Institut

fiir Kohlenforschung,

(Received

131 (1987)

S. GURTZGEN, D-4330

B. SPLIETHOFF Miilheim

and U. WILCZOK

an der Ruhr

(F. R.G.)

May 7, 1986)

Summary Bis(n3-allyl)nickel compounds react with catalytically prepared magnesium hydride (molar ratio, 1:2) in tetrahydrofuran at room temperature to give propene and finely divided (surface areas, about 100 m* g-l) amorphous solids. On heating (400 “C) followed by hydrogenation-dehydrogenation these are converted into crystalline Mg,Ni. Mg,Ni prepared by this method can be hydrogenated at 200 “C! under normal pressure. Bis(n3-allyl)nickel reacts with magnesium hydride (molar ratio, 1 :l) or with hydridomagnesiumchloride (molar ratio, 1:2) under the same conditions to give amorphous solids having the composition MgNiC,H,, x = 1.2 - 1.8, y = 2.3 - 3.4. At 690 “C!this solid is converted into a crystalline product which, on the basis of hydrogenation-dehydrogenation experiments and X-ray powder analysis, is suggested to be a mixture of Mg,Ni, the known carbide MgNi&, and MgO.

1. Introduction The present work originates from the discovery made in 1978 that commercial magnesium powder can be hydrogenated to magnesium hydride under mild conditions in an organic solvent (usually tetrahydrofuran (THF)) using soluble organ0 transition metal catalysts, in particular those of chromium and titanium [l]. This procedure allows the preparation on a large scale of a highly active pyrophoric magnesium hydride powder (or suspension) which has a high specific surface area (100 - 130 m* gg’) and which can be used both as a reagent for synthetic chemistry and as a high temperature hydrogen storage material [ 2, 31. When the catalytically prepared magnesium hydride is doped [4] with small amounts of bis(n3-allyl)nickel or bis(q3-allyl)palladium in THF or toluene it has been found that propene evolution occurs even below room temperature and that 70% or more of the ally1 groups originally present are liberated. An investigation of the stoichiometric reaction of *Paper presented at the International of Metal Hydrides V, Maubuisson, France, 0022-5088/87/$3.50

Symposium on the Properties May 25 - 30, 1986. 0 Elsevier

Sequoia/Printed

and Applications

in The Netherlands

164

catalytically prepared magnesium hydride or other metal hydrides with the bis(allyl)metal compounds showed surprisingly, in addition to propene evolution, the formation of intermetallic compounds or ternary hydrides in a highly active amorphous form [5]. In this connection it is of interest to note that commercial magnesium hydride reacts only very reluctantly with the bis(allyl)metaI compounds investigated, underlining the high reactivity of the catalytically prepared magnesium hydride [ 21. Here we report the reaction of the catalytically prepared magnesium hydride (1) (Scheme l), or of the soluble hydridomagnesium chloride (la) [l, 61, with bis(~3-allyl)nickel compounds (2a - 2c) which, depending on the molar ratio of the reactants, enables Mg,Ni (3a) or a m~esium nickel carbide (admixed with Mg~Ni) (4’) to be prepared under mild conditions.

BMgH, + R 1

+2Hz

- 2C3H4RR’ A RMg,NiH, _H THF

R

2a: R=R’=H 2b: R=CHs R’=H 2~: R=H, R’=CHJ

Mg,Ni _2H 2

3 (amorphous)

k&Nil& 2

3a (crystalline)

(crystalline)

2a - 2CsH6

1 + 2a

- 2C3H6 I

2HMg>z

MgTL

(amorphous)

la

W!&

+ Mg;&

+ N?O

(crystalline)

Scheme 1.

2. Experimental

procedures

All experiments were performed under argon. The elemental analyses performed in the labomto~es of Domis & Kolbe, M~heim an der Ruhr. An STAD/2/PL powder diffractometer (STOE) with a graphite monochromator (Cu Koli line) was used. Magnesium hydride (1) was prepared and isolated according to a previously published method [l] by hydrogenating magnesium powder (50 mesh) in THF at 20 “C (72 h) in the presence of a chromium catalyst; the resulting MgH, powder is of particularly high reactivity. The composition of 1 was as follows: 81.6% Mg, 7.0% H, 7.8% C, 2.8% Cl and 0.6% Cr. Hydrolysis yielded 5.5% THF and 0.4% n-butanol. A sample of 1 was heated in a thermovolumetric apparatus [7] from ambient temperature to 400 “C with a heating rate of 1 “C min-’ (standard conditions) and was found to contain 6.6 wt.% H, cont~inated by 1% but-l-ene and 0.2% butane. The specific surface area of the sample was 110 m2 g-l, were

165

Bis(allyl)metal procedures [ 81. 2.1. Preparation

compounds

were

prepared

and isolation of the amorphous

according

to

published

solids 3 and 4 (Table

I)

The solution of the appropriate bis(q3-allyl)nickel compound in THF was added dropwise at room temperature under stirring into a suspension of 1 in THF. The reaction mixture was stirred in a closed system for the time shown in Table 1. The volatile components were distilled off under vacuum (0.03 mbar) into two cold traps (-78 and -196 “C) in series. The gases condensed in the -196 “C trap were evaporated into a gas burette and analysed mass spectrometrically. The condensate in the -78 “C trap was analysed for diallyl and residual gases by gas chromatography. The residue from the distillation was triturated with fresh THF and the black pyropho~c solids 3 or 4 were separated from the solution by filtration through a D-4 glass frit, washed with THF (until the filtrate was colourless) and then pentane and dried at room temperature under high vacuum. Further data concerning the preparation of 3 and 4 and their analyses are shown in the Table 1. 2.2. Prepu~at~o~ of 4 and 4’ from 3 and .?a 2.30 g (15.9 mmol Ni) 3 and 1.90 g (13.5 mmol

Ni) ‘2a in 40 ml THF were stirred for 5 days at room temperature, resulting in the evolution of 75% of the calculated amount of propene and 6% of that of propane. By filtration and drying in high vacuum, 2.34 g 4 (78%; the composition 57.2% Ni, 23.5% Mg, 3.1% H and 14.2% C corresponds to Mga99Ni1.00H3tIsCl.al) were isolated as a black pyrophoric solid. According to the X-ray powder analysis 4 was amorphous. 0.68 g of the so-prepared 4 were heated to 400 “C in the thermovolumetric apparatus [7] (standard conditions), producing 677 ml gas (20 ‘C, 1 bar) having the composition 11% H,, 78% CH,+, 5% CsHs and 7% C4Hr0. The X-ray powder spectrum of the resulting solid 4’ exhibited diffuse reflections which could be assigned to MgNi3C, ]9,101. 3. Results and discussion The reaction of 1 with bis(~3-~lyl)nickel (Za) in the molar ratio 2:l in THF takes place with the evolution of propene and a little propane and is accompanied by the formation of a black solid having a specific surface area of 98 m2 gg’ and the approximate composition Mg,NiH, (3), apart from the organic material (15.2% C) (Table 1). (Part of the nickel remains in the solution and after evaporation of the THF a solid having the composition 8.9% Mg, 46.2% Ni, 35.7% C and 4.9% H was obtained.) Apart from very weak reflections of MgH,, the X-ray powder pattern indicated that 3 is amorphous. A sample of 11.30 g of 3 in a fully automated electronically controlled apparatus [ 111 was initially heated to 400 “C and then subjected to a series

166 TABLE

1

Reaction of the catalytically prepared magnesium hydrides with bis(??3-allyl)nickel compounds (2a - 2c) to the amorphous solids 3 and 4 in tetrahydrofuran at room temperaturea Experiment

Amount of lbb (g (mm+)

Amount of bis(q3-allyl)nickel compound (g (mmol))

Amount of THF (ml)

6.81 (223)

2a 15.76 (112)

250

2a 2.18 (20.0)

38

1.20 (39.9) 0.99 (33.0)

Reaction time (h)

Solid and Propene evolved q c amount (%I (g)

96’?

(313.70) 720gh

49

3 (1.80)

(2) 4j.k

66

3

2a 2.32 (16.5)

25

1.02 (33.7)

2b 2.60 (16.8)

35

1921

;z,

3

0.90 (29.9)

2c 2.51 (14.9)

38

144”

7o”

3

1.18 (35.2)

2a 1.18 (35.2)

50

116P

I’z”,

4

0.52 (15.3)

2a 4.33 (30.7)

40

96

:;I

4

4.86 ( 80.0)q

2a 5.63 (40.0)

123

48

(1.83)

(2)

(1.25) (1.20) (3.39) (1.70) 4 (2.02)

aUnless otherwise stated. b85% - 87% according to thermovolumetric analysis (see Section 2). CThe numerals in parentheses denote the percentage of propane (or isobutane) evolved. dWith respect to the bis(v’-allyl)niekel compound used. ‘?After 36 h the nickel concentration in the solution remained constant at 0.07 mol Ni I-’ ; after 96 h 2a was no longer detectable in the solution. fApart from very weak reflections of MgH2. EReaction temperature, 0 “C. hAfter 15 days 30% of 2a was unconverted; after 30 days only traces of 2a were detected in the solution. ‘After the thermal treatment (standard conditions; evolution of 300 - 350 ml gas g-‘;

of 34 hydrogenation-dehydrogenation cycles. Rapid heating of 3 (40 “C mir-l) above 120 “C (the internal temperature of the sample) led to exothermic gas evolution whereby the sample briefly reached 300 “C and 1200 ml of gas (20 *C, 1 bar) were evolved: the gas had the composition 65% H,, 5% CH,.,, 8% CaHs and 5% n-&HI0 and n-CeHs. Endothermic gas evolution occurred in the range 270 - 335 “C (internal temperature), resulting in the evolution of 1380 ml of gas (20 ‘C, 1 bar) having the composition 71% E-I,, 17% CH4, 2% CsHs and 8% n-C4Hs. Subsequent hydrogenation-dehydrogenation cycles indicated that the sample behaved kinetic~y in a manner similar to a sample of met~lu~ic~ly prepared MgzNi flZ]. After the first

167

Elemental

analysis

(%)

Empirical

formula

Yieldd (%)

Mg

Ni

H

C

Cl

35.9

40.3

3.9

15.2

0.4

40.6

34.4

5.0

16.2

Mg2.

38.4

39.0

4.5

15.5

44.3

35.0

4.9

9.5

47.3

32.0

6.1

22.7

54.3

20.1 20.5

82

M~z.l~Nil.~H5.68~1.83~10.02

- 83

Results powder

of X-ray analysis

Amorphous*

60

Amorphous**

Mg2.38Nil.&6.&l.w

74

Amorphousfqi

Mg3.osNil.dL3.2G.33

44

Amorphous*

17.8

Mg4.9Nil.ooH15.4C3.74

32

Amorphousf

3.2

17.1

Mgl.ooNil.,H3.41Cl.s2

89

Amorphous

51.6

2.9

18.7

Mgo.94Ni1.00H3.23C1.77

79

Amorphous

51.5

2.0

13.2

Mgo.96Ni1.~H2.2aC1.25C1o.2s

44

Amorphous

7.5

asNi ~.oo%

53%.

30

i

79% Hz, 21% CH4) the solid was identified as MgzNi by X-ray powder diffractometry. ‘Reaction temperature, 55 - 60 “C. kAfter 4 h 4% 2a was unconverted. ‘After 48 h 53% 2b was unconverted; after 8 days only traces of 2b were detected in the solution. mi-C4Hs. “After 4 days 19% 2c was unconverted. On-C4Hs. PAfter 48 h 60% C3H6 was detected; after a further 88 h 2a was no longer detectable in the solution. qHMgCl used instead ol MgH2.

five cycles (hydrogenation at 334 “C and 15 bar for 1.5 h; dehydrogenation at 334 “C and normal pressure for 1.5 h) the reversible H, capacity of the sample (hydrogenation at 200 and 260 “C and 1, 3, 5 and 10 bar for 1 h; dehydrogenation at 334 “C and normal pressure for 1.5 h) was constant at 2.65 wt.% (calculated value for Mg,NiH4, 3.62 wt.%) for the remaining 29 cycles. The comparatively low reversible H, content results from the presence of the inorganic components MgO and MgCl,: thermal decomposition of THF in the presence of 1 or of the “active magnesium” [2] leads to the formation of MgO together with n-butenes and n-butane, while MgCl* is a component of the catalyst used in the hydrogenation of the

168

magnesium [3, 131. After 34 cycles the sample (3a) was identical (X-ray powder pattern) with a sample of the met~lur~c~ly prepared Mg,Ni [12] and free of carbon; however, in contrast with the met~lurgic~ly prepared Mg,Ni, it can be hydrogenated at 200 “C under normal pressure. After repeated hydrogenation, Mg,NiH4 having the composition 42.9% Mg, 48.6% Ni, 2.64% H, 1.6% Cl and 0.2% Cr was obtained and identified by X-ray powder diffractometry [ 141. The reaction of 1 with 2a (molar ratio, 2:l; in THF) at 0 “C takes 30 days; however, at 55 - 60 “C the reaction is completed in 4 h. The employment of bis(~3-methallyl)nickel or bis(q3-crotyl)nickel (2b or 2c) instead of bis(q3-allyl)nickel (2a) at ambient temperature decreases the rate of reaction with 1 by a factor of roughly 2 and increases the amount of soluble nickel-containing reaction products; consequently the Mg:Ni ratio in 3 (Table 1, experiments 4 and 5) increases. The expe~ments using 2b and 2c also showed that according to X-ray powder diffractometry crystalline Mg,Ni (3a) is formed from 3 after one or two hydrogenation-dehydrogenation cycles (Table 2, experiments 4 and 5). When 1 is reacted with 2a in a molar ratio of 1:l instead of 2:1, propene is evolved and an amorphous solid (Fig. 1, spectrum a) with an Mg:Ni ratio of 1:l and 17.1% C (4, Table 1, experiment 6) precipitates. The solids with similar composition can also be prepared by reacting HMgCl [61 (la) with 2a in the molar ratio 2:l (experiment 8) or 1 and 2a in the molar ratio 1:2 (experiment 7): in the latter case the excess 2a is recovered unchanged. Thermal treatment (standard conditions~ of 4 (specific surface area, about 100 m2 g-i) produces mainly methane (Table 2, experiments 6 - 8) and the resulting solids exhibit only diffuse reflections in the X-ray powder diagram which can be assigned to the carbide phase MgNi3C, [ 9, lo]. It is also of interest to mention that MgzNiH, (3) can be converted into MgNiC, (4), and the latter thereafter into 4’, by reacting 3 with bis(q3-allyl)nickel (2a) (molar ratio, 1:l) in THF at room temperature (see Section 2). On further heating (4 h at 690 “C) the sample from experiment 6 is converted into a crystalline product (4’) whose X-ray powder pattern (Fig. 1, spectrum b), ignoring reflections arising from Mg,Ni and MgO, corresponds to that of the known ternary carbide MgNi$&s which has a cubic structure 19, lo]. A comparison of the powder diffraction data for MgNi3c a.,5 with those of MgNi3C, in 4’ (Fig. 1, spectrum b) is given in Table 3. Prolonged annealing of 4’ (5 days at 760 “C) causes the reflections for Mg,Ni in the X-ray powder pattern to disappear whereas those of the carbide phase remain unchanged (Fig. 1, spectrum c). In order to determine the amount of Mg,Ni present in MgNiC,, after the thermal treatment (standard conditions) 4 was subjected to a series of hydrogenation-dehydrogenation cycles under various hydrogenation conditions. Hydrogenation at 15 bar and 210 “C for 24 h followed by dehydrogenation led to the elimination of only minor amounts of H, and CH,, (Table 2, experiments 6 and 7) and the resulting solids exhibited only

2

3 3

4 4 4

4 5

6

136 133 114

321 482

(ml)

Vb

14 5 39

90 84

H,

67 86 43

‘I’ 12

CH~

1 -

.1

COHN

7 5 9

1

es8

(%)

treatment

Gas composition

Details of thermnl a

4 5 10

4 2

cala

and subsequent

MgNi&

MgNi&

treatment

e,f

e

Results of X-ray powder analysis after thermal

60 37

212d 202 78 55

(%)

22 45 27.5 24.0

65.0 60.0

Ni

1.25 0.9

0.9 1.7

H

6.1 3.4

1.2 3.8

C

Mg

Vb (ml)

Gas cornposition

Elemental composition after the hydrogenatio~-dehydrogenation cycle f%)

formula

Mg,,atj Nil.dh&zt

Mg,.,Nil.ooHl.12C0.46

Empirical

cycles for the solids 3 and 4

Details of hydrogenation-dehydrogenation cyclee

hydrogenation-dehydrogenation

MgNiG’

MgNW,

Mg,Ni Mg,Ni e

Results of X-ray powder analysis after the hydrogenation dehydrogenntion cycle

a0.5 - 2 g sample of the solid is heated to 400 “C at 1 “C min- 1 in the thermovolumetric apparatus [ 7 ] (standard conditions). bVolume of gas (20 “C; 1 bar) released per gram of sample during thermal treatment (dehydrogenation). CHydrogenation for experiments 4 and 5: normal pressure, 200 - 240 “C. Hydrogenation for experiments 6 and 7: 15 bar, 210 ‘C, 24 h. Dehydrogenation: standard conditions. dHydrogenation-dehydrogenation at normal pressure twice repeated. eDiffuse reflections. f3.2% C, 0.4% H.

8

I

Solid

Experiment

Details of the thermal treatment

TABLE

b)

a) 65

I

/

,

8

,

I

60

75

70

65

60

55

5

so

/

‘5

& LO

!

,

35

30

1 20

1

;:I 2 5

-

Fig. 1. X-ray diffraction patterns of MgNiC, before (spectrum a) and after heating to 400 “C (1 “C min-‘) followed by 4 h at 690 “C (spectrum b) and after annealing for 5 days at 760 “C (spectrum c).

diffuse reflections in the X-ray powder diagram, which could be assigned to the carbide phss6, MgNi&, [ 9, lo]. However, if after the thermal treatment 4 was hydrogenated at an elevated H, pressure (100 bar at 230 “C for 72 h), subsequent dehydrogenation (standard conditions) led to the elimination of the amount of H, (123 ml (g sample)-‘) expected for the presence of about 30 wt.% Mg,Ni in the sample. A second hydrogenationdehydrogenation cycle, under the same conditions, led to the liberation of the same amount of Hz. The thermovolumetric record [7] of the second dehydrogenation (Fig. 2) confirmed the presence of about 30 wt.% Mg,Ni in the sample and the X-ray powder pattern of the resulting solid 4’ (27.6% Mg, 59.8% Ni, 0.5% H and 8.5% C) indicated the presence of both the Mg,Ni phase and the MgNi&, phase. Only under very strong hydrogenation conditions (250 bar, 450 OC, 5 days) can 4’ (24.4% Mg, 58.0% Ni, 5.1% C, 0.6% H) be converted into a

171 TABLE 3 Comparison between X-ray powder diffraction [lo J and MgNi&, in 4’ (Fig. 1, spectrum b)

MgNi3C, a

MgNW’o.75

3.780 2.673 2.182 1.890 1.690 1.543 1.336 1.140 1.091 0.867

data of cubic (ao= 3.780 A) MgNi~C~.,~

hkl

I rel

d

I rel

(%I

@I

(%J

24 7 100 48 5 2 18 13 4 3

3.820 2.648 2.203 1.908 1.706 -

17 5 100 60 6 -

1.349 1.150 1.101 -

30 23 8

-

100 110

111 200 210 211 220 311 222 331

aFrom these data a cell dimension a0 = 3,816 .4 has been calculated 1 [:-cl

H2 [mil

I 350400450-

600 800 700 :

300-

500-

8' ,'

<'

i

I 9

12

15

18

9.

21

t Fig. 2. Thermovolumetric bar H2, 230 ‘C, 72 h): -, sample; m, AT.

IhJ

record of the dehydrogenation of hydrogenated MgNiC, (100 volume (millilitres) of gas evolved; - - -, temperature of the

carbon-free product (0.0% C, 0.4% H) whose X-ray powder pattern the presence of MgNi, [ 151, nickel [16] and MgO as expected decomposition of MgNi&, .

showed for the

172

On the basis of these results we suggest that 4’ is a mixture consisting of Mg,Ni (about 30%), MgNisC, and MgO. The crystallization and the hydrogenation-dehydrogenation rate of the Mg,Ni phase in 4’ seem to be adversely affected by the presence of carbon and/or carbide.

Acknowledgments The authors wish to thank Prof. I(. Yvon, Laboratoire de Cristallographie aux Rayons X, Universite de Geneve, Switzerland and Dr. D. NOI&IS, Arrhenius Laboratory, University of Stockholm, Sweden, for useful comments. We thank Dr. L. Schlapbach (Eidgen~ssische Technische Hochsc~~le Zurich) for a gift of Mg,Ni. References 1 B. Bogdanovi&, S. Liao, M. Schwickardi, P. Sikorsky and B. Spliethoff, Angew. Chem., 92 (1980) 845;Angew. Chem., Int. Edn. En@., 19 (1980) 818. B. Bogdanovifi, Eur. Patent 3564, 1982; Chem. Abstr., 91 (1979) no. 159787; U.S. Patent 4,554,253, 1985. 2 B. BogdanoviC, Angew. Chem., 97 (1985) 253; Angew. Chem., Int. Edn. Engi., 24 (1985) 262. 3 B. Bogdanovii: and B. Spliethoff, in T. N. Veziroglu, N. Getoff and P. Weinzierl (eds.), Hydrogen Energy Progress VI, Proc. 6th World hydrogen Energy Conf., Vienna, July 1986, Vol. 2, Pergamon, Oxford, 1986, pp. 797 - 810. 4 B. Bogdanovie, DE Offenlegungsschrift 3,247,365, 1984. U.S. Patent 4554,152, 1985. 5 B. Bogdanovii: and U. Wilczok, Patent applied. B. BogdanoviE, K.-H. Claus, S. Giirtzgen, B. Spliethoff and U. Wilczok, in preparation. 6 B. BogdanoviL: and M. Schwickardi, 2. Naturforsch., TeilB, 39 (1984) 1001. 7 B. Bogdanoviir and B. Spliethoff, Chem.-Zng.-Tech., 55 (1983) 156; manuscript 1074183. 8 G. Wilke, B. Bogdanoviti, P. Hardt, P. Heimbach, W. Keim, M. Kroner, W. Oberkirch, K. Tanaka, E. Steinriicke, D. Walter and H. Zimmermann, Angew. Chem., 78 (1966) 157;Angew. Chem., Int. Edn. Engl., 5 (1966) 151. B. Henc, P. W. Jolly, R. Salz, G. Wilke, R. Benn, E. G. Hoffmann, R. Mynott, G. Schroth, K. Seevogel, J. C. Sekutowski and C. Kruger, J. Organomet. Chem., I91 (1980) 425. 9 E. Scheil and L. Hiitter, 2. Metatlkd., 44 (1953) 387. L. Hiitter and II. H. Stadelmaier, Acta Metall., 6 (1958) 367. 10 Powder Diffraction File, Joint Committee on Powder Diffraction Standards, International Center for Diffraction Data, Swarthmore, PA, Card 28-624, 1978. 11 B. Bogdanovie and B. Spliethoff, Angew. Chem., 97 (1985) 269. 12 J. J. Reilly and R. H. Wiswall, Znorg. Chem., 7 (1968) 2254. J. Schefer, P. Fischer, W. Hiilg, F. Stucki, L. Schlapbach, J. J. Didisheim, K. Yvon and A. F. Andresen, J. Less-Common Met., 74 (1980) 65. 13 U. Westeppe, Dissertation, Ruhr-Universittit Bochum, 1985. 14 Z. Gavra, M. H. Mintz, G. Kimmel and Z. Hadari, Inorg. Chem., 18 (1979) 3595. 15 Powder Diffraction File, Joint Committee on Powder Diffraction Standards, Center for Diffraction Data, Swarthmore, PA, Card 3-1027, 1978. 16 Powder Diffraction File, Joint Committee on Powder Diffraction Standards, Center for Diffraction Data, Swarthmore, PA, Card 4-850, 1978.