Doping and pressure studies on YbBiPt

,


INTRODUCTION
Among Yb-based rare-earth (RE) compounds, the compound YbBiPt has attracted considerable attention due to its unusual heavy-fermion (H F) properties.The extraordinarily !arge Sommerfeld coefficient ( r-8 J K -2 mo1 -1 , below 400 mK) places YbBiPt as the largest effective mass HF compound known to datc. 1 Further, a cusp around 400 mK in the ac susceptibility 1 and a kink in the resistance at around 400 mK indicate that as with some sek'Cted Cerium HF compou nds, YbBiPt may be situated at the boundary bctween magnetic and HF states.Recent µ+SR experiments 2 in crushed powder samples of YbBiPt show evidence of spatially inhomogeneous magnetism below 500 mK.However, the question remains whether the observed inhomogeneous magnetism is caused by the sample quality (crushed powder) o r is an intrinsic property of YbBiPt.
With respect to the transport properties, recent resistance expcriments 3 on the series REBiPt indicate that a gradual evolution from semiconducting toward semimetallic behavior occurs as the RE varies from Nd to Yb. H all coefficient measurements on YbBiPt give a carrier count of about 0.04/formula unit, typical of a semimetal.
In view of the controversy involving t.he role of inhomogeneity and/or disorder in the physical interpretation of thermodynamic and transport properties of YbBiPt, in this paper we report on investigations of the effect of chemical ( Ybx Y 1 -.tBiPt) and hyd rostatic prtssure at low temperat ures on singlc crystals and crushed-powder samples.

II. MATERIALS AND EXPERIMENTAL TECHNIQUES
YbBiPt is face-ccntercd-cubic material with the MgAgAs structure. 4Single crystals of Ybx Y 1 _ _ JliPt (and YbzLu 1 _zBiPt) were grown out of Bi tlux 5 and characterized by powder x-ray diffraction.T he pressed-pellet samples of YbBiPt used for the heat-capacity e.xperiments under pressure were prepared by crushing single crysta ls, mi.xing them with a small amount of GE703 J varnish, and pressing at roughly 9 kbar at room temperature to obtain densities approaching 80% of the theoretical.The high-temperature electrical resistance under high hydrostatic pressure ( 1.5 K < T < 300 K, up to 16 kbar) has been measured using a four-probe technique with the sample placed in a piston-cylinder self-clamping berylliumcopper cell. 6At low temperatures (below 1 K ) the pressure cell was attached to the bottom of the mixing chamber of a dilution refrigerator.Heat-capacity measurements on single crystals of YbxYt -xBiPt (and YbzLu 1 _)JiPt) were carried out usirtg a quasiadiabatic thermal-relaxation tecbnique in the temperature range 1.5 K < T < 20 K. Finally, low-temperature heat-capacity experiments under hydrostatic pressure ( 300 rnK < T < 2 K, up to 8 kbar) were carried out at Lawrence Berkeley Laboratory. 7

III. RESUL TS AND DISCUSSION
Figure 1 (a) shows the high temperature ( 1.5 K < T < 300 K) resistance of a sing1e crystal of YbBiPt under hydrostatic pressure (to roughly 17.S kbar) .As the temperature decreases be1ow 300 K, the resistance decreases from the room temperature value, exhibiting an inftection point near 80 K ( T"').The temperature w here the inflection point occurs ( T 6.) remains almost unchanged as we apply pressure [inset of F ig. 1 (a)].l ndeed, inelastic neutron scattering experiments 8 in a polycrystallinc sample of YbBiPt reported a Jow-energy crystal-field excitation at approximately 6 me V, essen tially at the same order of magnitude as Ta• Association of the anomaly in t he resistance at T 6. with the crystal•field splitting is therefore rather suggestive.In other words, dT t.ldP, which is small, could be of the same order of magnitude as the pressuredependent Kondo-like scattering within this crystal-field doublet.9 Qualitatively, a very small (--0.5 K/kbar)  and negative dT 1:/dP is observed.We also believe that thermal expansion experiments will be a very important probe to investigate the crystal-field splitting of YbBiPt.
As the temperature is lowered below T ~• the resistance decreases, but with a slope close to half the slope found for T > Ta• At around 5 K a very fast drop in the resistance is observed.Our preliminary results of low-temperature resistance experiments under pressure are shown in Fig. 1 (b).The fast drop in R ( T) continues until another infiection point is observed at roughly 500 mK for tbe highpressure data [Fig. 1 ( b) shows the resistance at Iow temperature].The resistance at 16 kbar shows only a small variation when compared with the zero-pressure results.The interesting point is that the infiection point in the data collected at ambient pressure occurs at around the same temperature as the peak in ac susceptibility (Xac) in a single crystal. 1 Due to the small variation of R ( T,P) at low temperatures when compared with Xac the following question remains: ls the kink in the resistance a signature of a magnetic phase transition?
To investigate in more detail the effect of pressure on REBiPt materials, a series of resistance, magnetic susceptibility, and heat capacity experiments have been performed over a wide temperature range 10 on samples with substitution of Y or Lu for Yb.In Fig. 2  on YbxY 1 _xBiPt and YbzLu 1 _ )3iPt.Let us now emphasize the irnportance of the sample quality in the physics of YbBiPt.In Fig. 3 we show the low-temperature heat-capacity data for two different samples of YbBiPt.The first sample ( S 1) is a single crystal for which CJT at low temperatu res ( 40 mK < T < 400 mK) is r-8 J K'-2 mol-1 ; in addition, a well-defined maximum in the heat capacity near 400 mK is present. 1 The second sample (S2) was obtained by forming a pressed pellet from Lacerda et al. for the S2 sample, the shape and intensity of the transition are very different when compared to the S 1 results.For the high-temperaturc regime (above l K), SI and 82 present the same heat-capacity values.We want to point out that since YbBiPt seems to be strain sensitive, we also measured an annealed (750 °C for 10 days) single crystal of YbBiPt and the heat-capacity results were essentially similar to that föund for the Sl sample, indicating that there is no significant strain in the as-grown single crys tals.Very reccntly, similar strain dependence on C/T has been observed in the HF compound CeAI 3 • 12 Keeping the ba~ic differences between singlecrystalline and presscd-pellet heat-capacity results in mind, we now consider C measurements on a pressed pellet under pressure (sec Fig. 4).The heat capacity has been mea<>ure<l at 2, 6, and 8 kbar.The first striking result is the very small pressure depcndeuce of the heat capacity.The heat capac• YbBiPt, Slt2  ity at zero pressure (not shown) is only about 5% higher than the 2 kbar data.As we increase pressure, the maximum in C/T situated near 400 mK at zero pressure shifts to roughly 550 mK at 8 kbar and above 600 mK, Cis independent of pressure.Another HF compound with very low characteristic temperatures (below l K) and displaying a qualitatively similar pres:sure dependence of the heat capacity is CeA1 3 .For CeAl 3 , the C is most strougly pressure dependent for P < l kbar. 13For YbBil>t, further measurements of both heat capacity under pressure and thermal expansion on single crystals must bc d<Jne to try to elucidate the thermodynamics of this compound via scaling theories and a Grüneisen analysis.
To summarize, we have measured thermodynamic and transport properties of YbBiPt at low temperatures under hydrostatic pressure.The HF state of this compound seems to be unchangeable by a relatively !arge amount of nonmagnetic doping (Y or Lu).The heat-capacity exper~ iments under pressure in pressed-pellet samples reveal a small pressure dependence when compared with some selected HF compounds.Finally, the drastic difference between the heat-capacity signatures of single-crystal and pressed-pellet specimens clea rly indicates the importance of examining single-crystal samples when attempting to elucidate the unique physical propert.ies of YbDiPt.
FIG. 1.(a) Resistance as a function of tcmper--dture for a single crystal of YbBiPt.Tbc inset shows the pressure dependence of T ~ lsee the text}.(b) Low temperature resistance or YbBiPt at 16.5 kbar.

Figure 2 (
a) shows C/T as a function of temperature for some selected singlecrystal alloys.For pure YBiPt (and LuBiPt that is not shown) C!T at the Iowest temperature investigated ( 1.5 K) amounts to around 1 mJ K -2 mol-1 for both compounds.As we increase the Yb concentration to about 25% .[Fig.2(b)] a very ]arge C/T (comparable with YbBiPt) at very low temperatures (around 60 mK) is still observed.11The very !argeC/T for this reasonably large Y concentration is already an interesting result; this could mean that the !arge quasiparticle renorrnalization in YbBiPt ( the large Sommerfeld coefficient) is essentially unaffected by the lattice inhomogeneity and positive chemical pressure caused by Y substitutions.In Fig.2(b) we show the CJT (at 1.5 K) as a function ofYb content (x), as we •can see, only with a large amount of Y ( -50%) does the CJT at 1.5 K start to show considerable variations.Low-temperature heat-capacity experiments are currently in progress to study the phase diagram [Fig.2 ( b)] in greater detail.