A b s e n c e o f β - i r o n u p t o 4 4 GP a a n d 2 1 0 0 K : I n s i t u X - r a y O b s e r v a t i o n o f I r o n u s i n g K a w a i - t y p e A p p a r a t u s E q u i p p e d w i t h S i n t e r e d D i a m o n d Fig. 1 Fig. 1 Fig. 1. A cross-section of the octahedral specime n a ssembly. (i ) Pressure medium (MgO +5 wt % Cr 2 O 3 ), (i i) Thermal insulation sleeve (LaCr O 3 ), ( iii ) Cylindrical heater (Re), ( i v) Wedge-shape d electrode (Cu), (v ) D iamond powder, (vi) Sample sleeve (pressure s tandard) (MgO), (v ii ) Sample (M gO+Fe). A thermocouple (W97 / Re3-W75 / Re25, not s hown) is placed perpendicular to the pape r and in contact with the outer surface of the heater. The incident X-ray beam (arrow) was collimated to 0.05 mm horizontally and 0.1 mm vertically, whic h made it possible for us to acquire d iffractio n patterns from the sample and the pressure standar d independently . T he properties and states of iron at high-P/ T have been extensively studied by Earth scientists, l eading t hem to understand Earth’s core t hat is composed of iron. By the 1970s, four polymorphs of iron, namely α (bcc), γ ( fcc), ε (hcp) and δ (bcc), were k nown. In t he 1990s, however, based on experiments using a diamond anvil cell (DAC), a fifth phase called β was claimed to be present under conditions typically higher than 35 GPa and 1500 K. If the β phase really does exist, we might replace ε -iron as the most likely phase to constitute Earth’s inner core with β phase. However, both th e existence and structure of β phase are controversial; two different crystal structures have been proposed for β (double-hcp (dhcp) [1] and orthorhombic [2]) , while the absence of β up to 84 GPa and 3500 K has also been reported [3]. So far, exploration of β - iron has been impossible when using an eight cubi c anvil system (Kawai-type apparatus) equipped wit h tungsten carbide, because the maximum attainabl e p r e s s u r e w a s l i m i t e d t o l e s s t h a n 3 0 G P a . Nevertheless, by adopting sintered diamond (SD ) as the anvil material [4], it has become possible to generate pressure exceeding 40 GPa by using th e Kawai-type apparatus. U nder these circumstances , we studied t he phase relations of i ron using a Kawai-type apparatus equipped with SD , w hos e geometry aids c onsiderably in providing a quasi- hydrostatic environment for the sample . T he sample consisted of fine iron powder mixe d with MgO in a 1:1 weight ratio to s uppress grai n growth of the iron at high temperature. The sample was put into a semi-sintered MgO sleeve ( ) that served as the pressure standard. H igh pressure was achieved by compressing t he o ctahedral s pecimen a ssembly ( ) in a Kawai-type apparatus using “ SPEED 1500” press installed at beamlin e BL04B 1 [5]. E nergy dispersive X -ra y diffraction profiles for both the sample and the Mg O standard were collected along temperature cyclin g at fixed press loads of 660 and 800 tons at up to 1850 and 2100 K, respectively. P ressure wa s determined from the measured unit cell volume of the MgO pressure standard. 69 X - r a y 1 m m ( i ) ( i i ) ( i i i ) ( i v ) ( v i ) ( v ) ( v i i ) Fig. 2 Fig. 2(a) Fig. 2(b) Fig. 2. Examples of diffraction profiles of the sample. (a ) , (b ) , and (c ) are on the first cycle at 660 tons and (d ) is on the second cycle at 800 tons (see text). ε : ε -iron, γ : γ -iron, Fp: MgO mixed with the iron sample, B1; ( Fe, Mg)O with the B1 structure, R h: (Fe, Mg)O with the rhombohedral structure, asterisk : Fe 3 C. Numbers in parentheses are M iller i ndices. In addition to iron, diffraction peaks of MgO (mixed with the iron) were also observed . Fig. 2(c ) which Fig. 2(d) and Several diffraction profiles of t he sample ar e reproduced in . In the first cycle at 660 tons , iron initially assumed the ε phase at 37 GPa and 300 K, and it persisted at 1370 K and 38.4 GP a ( ) upon heating. Diffraction peaks of γ appeared at 1450 K and 38.4 GPa and intensifie d s imultaneously with r eduction of ε as t he temperatur e increased to 1850 K and 39.2 GPa ( ). A substantial amount of ε still survived at 1850 K due to t he t emperature gradient through t he s ample. U pon c ooling, enhancement of peaks for ε wa s noticed with a simultaneous reduction of γ at 1350 K and 37.1 GPa, but a certain amount of t he γ phase remained even after quenching to 300 K and 34 GPa ( ). In the second cycle at 800 tons , no change was observed up to 1700 K. At 1750 K and 41.9 GPa, t he growth of γ was recognize d together with a reduction of ε , was rapid wit h increasing temperature up to 2100 K and 44 GP a ( ). The mutual growth relations of the γ and ε phases corresponding to heating and c oolin g completely deny t he presence of any phas e between the stability fields of γ ε up to 44 GPa. Moreover, we c ould not find any c haracteristic 70 * * * * * * * * * * * F p ( 2 2 0 ) ε ( 2 0 0 ) B 1 ( 2 0 0 ) F p ( 1 1 1 ) ε ( 1 0 0 ) ε ( 1 0 1 ) ε ( 1 0 2 ) F p ( 2 2 0 ) ε ( 1 1 0 ) + F p ( 3 1 1 ) B 1 F p γ + F p ε B 1 F p B 1 ε + F p ε ε γ ε B 1 γ R h ( 0 0 3 , 1 0 1 ) F p γ + F p ε γ R h ( 1 0 2 ) ε ε γ R h ( 1 0 4 , 1 1 0 ) ε + F p R h ( 1 0 5 , 1 1 3 , 2 0 1 ) ε + F p F p ε B 1 ( 1 1 1 ) F p B 1 ε + F p ε ε B 1 ( 2 2 0 ) γ ( 2 0 0 ) B 1 ( 3 1 1 ) γ ( 2 2 0 ) γ ( 1 1 1 ) + F p ε ( d ) 4 4 . 0 G P a , 2 1 0 0 K ( c ) q u e n c h e d t o c a . 3 4 G P a , 3 0 0 K ( b ) 3 9 . 2 G P a , 1 8 4 5 K ( a ) 3 8 . 4 G P a , 1 3 7 3 K 5 0 6 0 7 0 8 0 9 0 1 0 0 E n e r g y ( k e V ) [ 2 θ = 6 . 0 1 8 d e g ] I n t e n s i t y of triple Fig. 3. Summary of the experimental results. The solid and open circles denote the conditions at whic h the growth of γ and ε phases were observed, respectively. The thick line s hows the ε - γ boundar y constrained in the present study. The open squares and diamonds are the conditions under which β phases of dhcp and orthorhombic structures were reported, respectively [1, 2]. The solid square indicate s α - γ - ε point [6]. The dashed lines are the previously reported melting curve and ε - γ boundary iron [3]. Bars indicate error in temperature and pressure. N ote that we observed growth of the γ phas e at the P/T conditions where β phase has been reported as a stable phase in previous studies [1, 2] . peaks of dhcp- and orthorhombic iron in any of th e diffraction profiles. All the identified phases of iron are plotted on a P/T s pace along with selecte d previous results ( Fig. 3 ). As t he β phase(s) has been observed in DAC without using pressure medium [1] or using high-strength pressure medi a [2], strong uniaxial stress on t he sample might cause an incomplete phase transition from ε to a “metastable” β phase. References [1] S. K. Saxena and L. S. Dubrovinsky, Am . Mineral. 85 (2000) 372. [2] D. Andraul t et al . , Am. Mineral. 85 (2000) 364. [3] G. Shen et al . , G eophys. Res. Lett. 25 (1998) 373. [4] E. Ito et al . , Geophys. Res. Lett. 25 (1998) 821. [5] W. Utsumi et al . , S pec. Issue Rev. High Press. Sci. Technol . 6 (1997) 512. [6] T. Uchida et al . , J. Geophys. Res. 106 (2001) 21799. [7] Atsush i K ubo, Eiji Ito , Tomo o Katsura, Toru S hinmei, Hitosh i Y amada, Osamu Nishikawa, Maoshuang Song and Kenichi F unakoshi, Geophys. Res. Lett. 30 (2003) 1126 . Atsush i Kubo a , Eiji Ito a and Kenich i Funakoshi b (a) Okayama Universit y (b) SPring-8 / JASR I E-mail: akubo @m isasa.okayama-u.ac.j p 71 L i q u i d 0 1 0 2 0 3 0 4 0 5 0 6 0 7 0 8 0 T e m p e r a t u r e ( K ) P r e s s u r e ( G P a ) 5 0 0 1 0 0 0 1 5 0 0 2 0 0 0 2 5 0 0 ε δ γ α F e
home
- JP
- EN