Arxiv:1702.07144V3 [Cond-Mat.Str-El] 23 Jul 2017 Rect Measurements of Thermal Diﬀusivity[15] Have Pro- Fig

Metallicity without quasi-particles in room-temperature strontium titanate

Xiao Lin1, Carl Willem Rischau1, Lisa Buchauer1, Alexandre Jaoui1, BenoˆıtFauqué1,2 and Kamran Behnia1 (1) Laboratoire Physique et Etude de Matériaux(CNRS-UPMC), ESPCI Paris, PSL Research University, 75005 Paris, France (2) JEIP, USR 3573 CNRS, Collègede France, PSL Research University, 75005 Paris, France

(Dated: April 28, 2017) Cooling oxygen-deficient strontium titanate to liquid-helium temperature leads to a decrease in its electrical resistivity by several orders of magnitude. The temperature dependence of resistivity follows a rough T3 behavior before becoming T2 in the low-temperature limit, as expected in a Fermi liquid. Here, we show that the roughly cubic resistivity above 100K corresponds to a regime where the quasi-particle mean-free-path is shorter than the electron wave-length and the interatomic distance. These criteria define the Mott-Ioffe-Regel limit. Exceeding this limit is the hallmark of strange metallicity, which occurs in strontium titanate well below room temperature, in contrast to other perovskytes. We argue that the T3-resistivity cannot be accounted for by electron-phonon scattering àla Bloch-Gruneisen and consider an alternative scheme based on Landauer transmission between individual dopants hosting large polarons. We find a scaling relationship between carrier mobility, the electric permittivity and the frequency of transverse optical soft mode in this temperature range. Providing an account of this observation emerges as a challenge to theory.

I. INTRODUCTION cubic[17, 22]. In this regime, the mean-free-path of carriers becomes shorter than their wavelength. At room The existence of well-defined quasi-particles is taken temperature, it falls well below the interatomic distance, for granted in the Boltzmann-Drude picture of electronic the lowest conceivable length scale. Strontium titanate transport. In this picture, carriers of charge or energy is the strangest known metal according to the criteria for are scattered after traveling a finite distance. The Mott- strangeness widely used. Since quasi-particles in this sys- Ioffe-Regel(MIR) limit is attained when the mean-free- tem are not well-defined in a time scale long enough to path of a carrier falls below its Fermi wavelength or the allow the existence of a scattering time, charge transport interatomic distance[1]. In most metals, resistivity sat- cannot be reasonably described by a phonon-scattering urates when this limit is approached. But in “bad”[2] picture. This was the approach in previous attempts[16– or “strange”[3] metals, it continues to increase[4]. It is 19, 24, 25, 27] to understand charge transport. We argue indeed strange when the mean-free-path persists to fall that metallicity is caused by the temperature dependence after attaining its shortest conceivable magnitude and of- of the transmission coefficient along Landauer canals be- ten unexpected behavior is considered bad. tween dopants. Such a picture does not require a scattering time or a mean-free-path. We find an empirical link In most cases, bad metals present a linear tempera- between mobility and permittivity, which gives a reason- ture dependence beyond the MIR limit. Bruin et al.[5] able account of the experimental data. Providing a quan- recently noticed that the T-linear scattering rate in nu- titative account of this observation emerges as a specific merous metals, either ordinary or strange, has a similar challenge to transport theory. magnitude. This observation suggests the relevance of a universal Planck timescale (τP ∼ ~ , where kB is the kB T Boltzmann and ~ the reduced Planck constants) to electronic dissipation and motivated a theoretical attempt II. RESULTS to quantify an upper limit to the diffusion constant in incoherent metals[6]. During the past years, several the- A. Metallicity beyond the MIR limit in strontium ories of charge transport with no need for well-defined titanate quasi-particles were proposed[7–14]. More recently, di- arXiv:1702.07144v3 [cond-mat.str-el] 23 Jul 2017 rect measurements of thermal diffusivity[15] have pro- Fig. 1a shows the temperature dependence of re- vided additional evidence for quasi-particle-free thermal sistivity in SrTiO3−δ as the carrier concentration is transport at room temperature in cuprates. changed from 2.4 1017cm−3 to 3.5 1019cm−3. At low We show here that n-doped SrTiO3 is an unno- temperature (T 100K), the system displays a T- ticed case of strange or bad metallicity. The resistiv- square resistivity[21, 22] as expected for electron-electron ity of this dilute metal decreases by orders of magni- scattering[28] in a Fermi liquid. Plotting resistivity as tude upon cooling to cryogenic temperatures[16–20]. In a function of T2 (see Fig. 2 in ref.[22]), one can see the low-temperature limit, resistivity follows a T-square that inelastic resistivity becomes T-square in the low- behavior[21, 22], as expected in a Fermi liquid. Above temperature limit. A wide range of dense Fermi liquids 100 K, however, the temperature dependence is close to display Kadowaki-Woods scaling[29] of their T-square TABLE I: How the MIR limit of metallicity is exceeded in different perovskytes- The table compares the Fermi wave-vector, the resistivity and the mean-free-path at 400 K in four different systems using equation 1. The table lists an underdoped, an optially-doped cuprate and strontium ruthenate and compares them with doped strontium titanate. In the case of YBa2Cu4O8, we assume that there are four Fermi surface pockets with a kF corresponding to the square root of the frequency of quantum oscillations. In contrast to other systems, the MIR limit in doped strontium titanate occurs well below room temperature. −1 Compound kF (nm ) ρ(400K) (mΩcm) `(400K)(nm) kF ` (400K) TkF `=1 (K) Reference La1.72Sr0.18Cu04 6.6 0.32 0.35 2.2 850 [4, 41] YBa2Cu4O8 1.4 0.36 0.8 1.1 450 [42, 43] Sr2RuO4 5.6 0.2 0.38 2.1 800 [44] 17 −3 SrTiO3−δ (n=4 10 cm ) 0.24 5900 0.2 0.05 100 This work

resistivity prefactor and their electronic speciﬁc heat. 0 Since the latter is proportional to carrier concentration, 1 0 S r T i O 3−δ the case of low-density systems such as doped SrTiO3 1 0 - 1

) is diﬀerent[30]. However, the scaling between the resis- m

tivity prefactor and Fermi energy persists. Indeed, the c - 2

Ω 1 0

( magnitude of this prefactor decreases by several orders

ρ of magnitude with changing carrier concentration[22]. 1 0 - 3 Fig. 1b shows the temperature dependence of the ex-

- 4 a ) 1 1 0 tracted Hall mobility, µ = neρ . Above 100 K, it displays 4 1 0 a temperature dependence close to T−3 and barely varies

) - 3 s

. T with carrier concentration. Our data is in agreement with

V 3 (

/ 1 0 the less extensive data reported previously by Tufte and 2 1 7 - 3

m n ( 1 0 c m ) = Chapman[17]. Fig. 1c shows how the exponent of inelas- c

( 2 2 . 4 4 . 1 8 . 0 tic resistivity evolves as a function of temperature. The y 1 0 t

i α

l 1 2 1 4 2 6 i total resistivity can be expressed as ρ = ρ0 + AT . As

b 3 2 5 4 8 3

o seen in Fig. 1c, α is 2 at very low-temperature (T < 50K) 1 1 7 0 2 9 0 m 1 0 b ) but smoothly rises with warming. One does not expect the T-square resistivity to survive at the Fermi degener- 1 1 0 T ( K ) 1 0 0 3 0 0 acy temperature. Therefore, the electron-electron scat- α tering origin of the T-square term in the dilute limit has 3 . 0 ρ= ρ + T 0 been put into question[31, 32]. We note, however, that the smooth variation of α, makes the detection of the up- 2 . 5 per boundary of T-square resistivity impossible. As one

α can see in the ﬁgure, α peaks around 150 K at a value T slightly exceeding 3 and then decreases with increasing 2 . 0 A F D temperature. c ) In the zero temperature limit, the magnitude of elec- 0 5 0 1 0 0 1 5 0 2 0 0 2 5 0 3 0 0 T ( K ) tron mobility, set by defect scattering can become remarkably large in doped strontium titanate[33]. As a

FIG. 1: Resistivity and mobility in doped SrTiO3: a) consequence, quantum oscillations are easily detectable Temperature dependence of resistivity in SrTiO3−δ as the car- in moderate magnetic fields[34–36]. The large magni- rier concentrations is tuned from 1017 to 1019 cm−3. Room- tude of defect-limited mobility at low temperature and temperature resistivity is several hundred times higher than its evolution with carrier concentration can be explained low-temperature resistivity. b) Hall mobility as a function of in a simple model[37] invoking the long effective Bohr temperature. Above 100 K, the mobility does not depend on radius of the quantum paraelectric[38] parent insulator. carrier concentration and is roughly cubic in temperature. c) The focus of the present paper is charge transport Assuming that inelastic resistivity follows a power law, i.e. α above 100 K where mobility presents a strong temper- ρ = ρ0 + AT , one can extract the exponent, α by taking a ature dependence and almost no variation with carrier logarithmic derivative: α = dln(ρ − ρ0)/dlnT , as in the case of cuprates[39]. At low temperature, α ' 2. Above 50K, it concentration over two orders of magnitude (Fig. 1b). starts a significant shift upward and exceeds 3 around 150K, Early studies[16–19] attributed the T−3 variation of mo- before steadily decreasing afterwards. The antiferrodistortive bility to the scattering of carriers by phonons. There was transition[40] is the source of the small anomaly at 105K. a controversy between Fredrikse et al.[18] (who invoked longitudinal optical phonons) and Wemple et al.[19] (who argued in favor of transverse optical phonons). Neither during this early debate[16–19], nor in more recent the- SrTiO 17 -3 è. Namely: 3- n=3.7 10 cm 300 2 2 MFP (Eq.1) ne τ 1 e 2 σ = = kF è (1) 100 MFP (non-degenerate) m∗ 3π2 ~ 

-1/3 The Fermi surface of n-doped SrTiO3 in the dilute n limit is known. It is located at the center of the Bril- 10 -1 k F louin zone, because a band originating from Ti atoms’ t orbitals has a minimum at the Γ−point[21, 35, 40].

(nm) e l The threefold degeneracy of this band is lifted by tetrag- 1 onal distortion and spin-orbit coupling. As the system is a doped, the three bands are filled successively and three concentric Fermi surfaces will emerge and grow in size 0.1 one after the other. This is a picture based on band Degenerate Non-degenerate 0.04 calculations[21] and confirmed by experimental studies 2 10 100 400 18 −3 a) T (K) of quantum oscillations[34–36]. Below 10 cm , the first critical doping for the emergence of secondary pockets, there is a single Fermi surface with moderate (1.6) anisotropy[35, 36, 40]. Therefore, kF can be quantified in a straightforward manner. Fig. 2a shows the temperature dependence of the mean-free-path extracted from Eq. 1 in a dilute sample whose resistivity was measured up to 400 K. The carrier density is the one given by the Hall coefficient. As previously reported in ref. [17] , the Hall coefficient does not display any notable temperature dependence. Moreover, the carrier density according to the magnitude of the Hall coefficient matches the Fermi surface volume according to the frequency of quantum oscillations[36]. Therefore, there is little ambiguity re- garding carrier concentration. b) As seen in Fig. 2a, the mean-free-path becomes shorter −1 than the inverse of the Fermi wave-vector kF and then the lattice parameter a. Now, at this carrier concentra- ∗ FIG. 2: Breakdown of the Mott-Ioffe-Regel limit: tion, with an effective mass of m =1.8me, the Fermion a) Temperature dependence of the mean-free-path in dilute degeneracy temperature is as low as 15 K[34]. As one sees metallic SrTiO3−δ sample up to 400 K. Blue circles repre- in the figure, this is the temperature at which de Broglie √ h sent è, extracted from resistivity using Eq.1. Red diamonds thermal wavelength, Λ = ∗ becomes compara- 2πm kB T represent the mean-free-path in the non-degenerate regime: −1/3 −1/3 ble to the interelectron distance n . Above this tem- `0 = ` n . Also shown are the de Broglie thermal wave- e e Λ perature, the electrons become non-degenerate and fol- −1/3 length, Λ, the interelectron distance n , the inverse of the low a Maxwell-Boltzmann distribution. As a consequence Fermi wavelength (k−1) and the lattice parameter, a=0.39 F and thanks to thermal excitation, the typical carrier mo- nm. The system remains metallic even at 400 K, in spite of mentum is larger than k . In other words, the typical exceeding the MIR limit by all conceivable criteria. b) A color ~ F 0 velocity is not the Fermi velocity, but the thermal ve- plot of è, extracted from the resistivity data of Fig. 1 in the q (T, n) plane. Different crossovers are shown. The MIR limit 2kB T locity, vT = m∗ . Let us define the mean-free-path of is exceeded in a high-temperature window narrowing down 0 −1/3 electrons with thermal velocity as è = τvT = èn /Λ. with increasing concentration. 0 As seen in the figure, because of è > è, the crossover temperatures shift upward. However, the picture remains qualitatively the same. In the non-degenerate regime, the relevant wavelength is Λ and not the Fermi wavelength. oretical accounts of the temperature dependence of mo- As seen in the figure, the mean-free-path becomes shorter bility in n-doped SrTiO3[23–27], the magnitude of the than Λ too. mean-free-path was discussed. A room-temperature resistivity of 2.4 mΩ.cm implies a scattering time of τ ' 6fs. This is shorter than the Planck time[5] (τP = ~ ) at room temperature, In the Drude-Boltzmann picture, the conductivity, σ, kB T of a metal depends on fundamental constants and a num- τP (300K) = 25.3fs. This is the typical time it takes for ber of material-related properties. These are either, the a carrier with a thermal velocity, vT , to move as far as ∗ Λ scattering time τ, and the effective mass, m , or the its de Broglie thermal wavelength ( = 1.77τP ). By vT Fermi wave-vector, kF and the electron mean-free-path, all accounts, room-temperature quasi-particles in dilute

B. Comparison with graphene and metallic silicon R i n G r a p h e n e s q ( n 1 / 2 = 3 . 5 1 0 6 c m - 1 ) 1 0 0 0 0 2 D R = ρ/ λ i n S r T i O F 3 - δ The temperature dependence of resistivity above 100 ) ( n 1 / 3 = 3 . 3 1 0 6 c m - 1 ) Ω 3 D

( K and the magnitude of inelastic resistivity are very dif-

R 1 0 0 0 ferent from dilute metallic systems subject to phonon scattering. Let us compare strontium titanate with two exemplary systems. A recent study of charge transport

1 0 0 in graphene[45] has shown the relevance of the Bloch- Gr¨uneisenpicture of electron-phonon scattering in our 1 0 0 S i : P ( n = 6 . 2 1 0 1 8 c m - 3 ) temperature window of interest. In this picture, above 1 8 - 3 S r T i O δ( n = 5 . 4 1 0 c m ) ) 3 - a characteristic temperature (usually, but not always,

m 1 0 c

the Debye temperature) resistivity displays a linear tem- Ω

m perature dependence. This happens because all phonon ( 1 ρ modes capable of scattering electrons are thermally pop-

ulated and the phase space for scattering is proportional 0 . 1 to the thermal width near the Fermi level. Far below 3 1 8 - 3 1 0 S r T i O δ ( n = 1 . 4 1 0 c m ) 3 - the characteristic temperature, the resistivity is expected P b T e ( p = 1 . 1 1 0 1 8 c m - 3 ) 2 5 4 ) 1 7 - 3 to display a much stronger T (in 3D) or T (in 2D) 1 0 K T a O δ ( n = 6 1 0 c m ) ( W e m p l e )

m 3 - c 1 temperature dependence because the typical wave-vector Ω 1 0

( of the thermally excited phonons becomes smaller with

ρ 1 0 0 decreasing temperature. As a consequence, the capac-

- 1 ity of scattering phonons to decay the momentum flow 1 0 decays rapidly with temperature. Efetov and Kim[45] 1 1 0 1 0 0 showed that this picture gives a very satisfactory ac- T ( K ) count of inelastic resistivity in graphene as the carrier density is tuned from 1.36 to 10.8 1013cm−2. Let us note that in gated graphene, inelastic (that is, temperature- FIG. 3: Comparison with other dilute metallic sys- dependent) resistivity is far from dominating the total tems: a) R of a one Fermi-wavelength-thick sheet of resistivity. This is seen in Fig. 3a, which compares the SrTiO compared with graphene[45] with comparable areal 3−δ temperature-dependence of resistivity in n-doped SrTiO3 carrier density. b) Temperature-dependence of resistivity in and graphene at an identical interelectron distance. As SrTiO3−δ and in metallic silicon[46] with comparable volume one can see in the figure, resistivity changes by 200 in carrier concentration. c) Temperature-dependence of resis- SrTiO3 and by 1.6 in graphene. Charge conductivity is tivity in SrTiO3−δ, metallic PbTe and metallic KTaO3−δ[47] in similar range of carrier concentration. In silicon and higher in graphene by a factor of two at low temperature graphene, the change in resistance caused by phonon scat- and by two orders of magnitude at room temperature. A tering of electrons is a modest part of the total resistance. In comparison with the case of metallic silicon[46] reveals a all other systems, which are close to a ferroelectric instability, similar discrepancy. As one can see in Fig. 3b, cooling resistivity changes by several orders of magnitude. phosphorous-doped silicon leads to a modest change in resistivity, much less than what is seen in doped SrTiO3. As a consequence, while at low-temperature, strontium titanate has a significantly higher mobility, the opposite metallic strontium titanate do not live long enough to is true at room temperature. In contrast to strontium qualify for a scattering-based picture. titanate, in both graphene and silicon, one can use Eq. A color plot of the mean-free-path in the (T, n) plane 1 one to extract a meaningful mean-free-path, which in- produced from the data of Fig.1 and Eq. 1 is presented creases by a factor of two or less as the system is cooled in Fig. 2b. The MIR limit is exceeded in a high- down from room temperature to helium liquefaction tem- temperature window, widening with the decrease in car- perature. rier concentration. The hitherto unnoticed specificity of It is also instructive to compare SrTiO3 with two SrTiO3 is that it becomes a strange metal well below other dilute metals close to a ferroelectric instability. room temperature. The data in Table 1 gives an idea of As one can see in Fig.3c, both PbTe and KTaO3 show the “strangeness” or the “badness” of SrTiO3 compared comparably large changes in their resistivity upon cool- to two other perovskytes. As seen in the table, at 400 K, ing. Reports on KTaO3[19, 47] and on lead chalcogenide kF ` in this system is much lower than in cuprates or in salts (PbTe, PbSe and PbS)[48, 49] indicate that high- Sr2RuO4. Contrary to the two other systems, the passage temperature mobility does not vary with carrier con- beyond the MIR limit occurs well below room tempera- centration, but shows a strong temperature dependence ture implying the necessity of a transport picture with (µ ∝ T −λ, with λ ' 2.5 − 3). Both these features are no reference to well-defined quasi-particles. This is the similar to what we see in strontium titanate. Only in first principal message of this paper. strontium titanate, however, the magnitude of the room

0 5 3 ) , found by 3 0 - 4 1 e . . 6 4 7 m 2 4 8 2 5 1 2

B c a 7 Λ Λ 0 1 0 a) Temperature 3 0 / 3 3 / 1 1 1 - - ( ) )

3 =1.8m 3 - - n ∗ m m c c 9 7 1 1 0 0 0 5 , the distance between 1 1 2

3 7 4 ) . . − 1 K 2 (

( ( δ T −

cm 3 0 19 O 0 i 2 T 10 r S × 7 . 0 5 1 to 1 3 ) ) − b a 0 . 0 2 1 cm

2 5 0 0 0 0 8 6 4 5 0 5 0

4 2 1 2 2 1 0 1

) m n ( h t g n e ) m n ( G n /

l 17 Length scales and transmission σ 2 10 × In this second picture, conductivity is set by trans- Let us now consider different length scales of the sys- 4 . quantum oscillations at lowradius temperature. was The calculated effectivedata Bohr using reported this in massfrom [38]. resistivity and data b) the forpared Transmission different permittivity to carrier coefficient Λ concentrations extracted com- ging its own hole’[50]laron in picture a has polar been semiconductor.strontium titanate[56, frequently The 57]. invoked po- in the case of mission among dopantwell-defined sites quasi-particles. andwith Can there decreasing conductivity temperature is increase erate in no such a a need metallic pictureincreases for behavior? and the gen- transmission Thisdopant probability can between sites. adjacent happenpermittivity if In increases cooling the with caselower cooling[38]. of the strontium temperature, Therefore,tial. titanate, the As the static shallower a consequence, thedecreasing polarons temperature donor become and less poten- conductivity bound is with metallic. tem. Fig.of interest where 5arier mobility shows concentration. becomes them independent2 When in of carrier car- our density temperature changes range from FIG. 5: dependence of differentdensities. length scales at Theculated two de different assuming Broglie carrier an thermal effective wavelength, mass Λ, of was m cal- electrons varies between 16 to 4 nm. Taking the low-

by by transmission

Driven b)

Decreasing temperature

e - e e and other metals close to a ferro-

semiconductor 3 III. DISCUSSION scattering by by

a) In a Drude-Boltzmann picture of transport, riven behavior at low temperature). On the other D a) 5 Two pictures of temperature-dependent con-

- -

- e e or T e To sum up, in silicon and graphene, the temperature- Fig. 4 sketches two alternative descriptions of charge A. Conduction as transmission in a doped polar 4 FIG. 4: induced change in resistivitymean-free-path implies and a modest follows change(a a in slow Bloch-Gruneisen T-linear behavior behaviorT at high temperature and a fast ductivity: temperature mobility is lowstrange enough metal to below make room it temperature. a clearly cooling the system leadsConductivity to a increases rarefication of becausespatial scattering events. distance) the between two timecreases successive scattering interval with events (and cooling. in- quasi-particle the across Such the distance a ining time picture and two requires in scattering space a separat- events.port, well-defined b) each In dopant aconnected (each Landauer oxygen to picture vacancy of other here)nels. trans- adjacent is Conductivity reservoirs increases a by uponmission reservoir conducting coefficient cooling enhances. because chan- the Suchwell-defined trans- a quasi-particles. picture does not require hand, doped SrTiO electric instability doapparent not mean-free-path of fit carriersmagnitude. within changes by this It orders becomes of picture.and very in short The at thephysically high case meaningful. temperature, of strontium titanate, too short to be conduction. In the semi-classictransport picture (Fig. of metallic 4a), charge conductivitying increases temperature, with because decreas- the the momentum scattering events ofcome relaxing a less charge-carrying frequentconsider with quasi-particle an decreasing be- alternative temperature.dopant is picture hosting Now, (Fig. a polaron, ‘an 4b), electron trapped where by dig- each ∗ temperature effective mass of m =1.8me seen by quan- B. Mobility, diffusivity and their link to electric tum oscillations[34], one finds that the de Broglie ther- permittivity mal wave-length is Λ ' 3 − 6nm in this temperature range. Thus, the system is mostly non-degenerate. Elec- In a non-degenerate semiconductor, mobility and diffu- tric permittivity, , is large ((300K) ' 300)[38, 53] and sion constant, D, of carriers are linked through Einstein’s increases steadily with cooling. As a consequence, the equation[54]: 2 ∗ 4π~ effective Bohr radius (aB = m∗e2 ) varies between 8 and 40 nm. This puts the system above the Mott criterion for 1/3 ∗ µkBT metallicity (n aB > 0.25[51]), which, one shall not for- D = (3) get, is relevant to degenerate electrons[52]. These length e scales are in the same range of magnitude. They are all The temperature dependence of D extracted from the longer than the lattice parameter (0.39nm). The question mobility data of the sample with a carrier concentration is how to picture conduction as transmission[55] along a of 2.4 1017 cm−3 is shown in Fig. 6a. Note that since mo- random network of donors. bility does not show any detectable variation with carrier concentration, this curve is representative of D for car- Since mobility is found to be independent of carrier rier concentrations in the range of 1017-1019cm−3. The concentration above 100K, we can express conductivity magnitude of D obtained in this way is remarkably low, as: well below 1cm2/s. For comparison, the diffusion constant of electrons in silicon rises from 35 cm2/s at room temperature to ∼100 cm2/s at 100 K[58]. Hartnoll recently proposed a lower bound to diffusion −1 constant in incoherent metals[6]: σ = ρ = G0n < T > (2)

2 ~vF DH = (4) 2e2 kBT Here, G0 = h , is the quantum of conductance and < T > represents the average transmission between a As one can see in Fig. 6a, the experimental D decreases dopant site and its immediate neighbors. Note that in faster than T−1. Hartnoll boundary was obtained for de- three dimensions, < T > is proportional to mobility (µ = generate systems. Ours is non-degenerate and the typical e 2 h < T >). velocity is thermal velocity and not vF . The Fermi velocity changes drastically with carrier density, but the The temperature dependence of < T >, extracted from experimental D does not. As defined in Eq. 4 ,D , does the experimental data using Eq. 2 is shown in Fig. 5b. H not appear to be directly relevant to our observation. As seen in the figure, a hundred-fold enhancement in car- An obvious piece of the puzzle of charge transport in rier concentration does not affect < T >. One would strontium titanate is the fact that the insulating parent have expected that the Landauer transmission between is a quantum paraelectric with a strongly temperature- one site and its more or less distant neighbors would de- dependent electric permittivity[38, 53]. We have found pend on the length of the trajectory and the details of that in our temperature range of interest D is propor- the structural landscape. It is therefore surprising that tional to 1.5. This empirical observation is the second the averaged transmission coefficient, < T >, is insen- principal message of this paper. sitive to such a large change in the average interdopant The scaling between static electric permittivity, , and distance. 0 the frequency of the Transverse Optic (TO) soft mode[59, Fig. 5b also compares the magnitude of < T > with 60], ωTO in strontium titanate is well-known.√ Yamada 2 −1 Λ . the latter represents the areal uncertainty in the and Shirane[60] showed that ωTO = 194.4 within spatial localisation of a non-degenerate electron in three experimental resolution. In a polar crystal, there is a dimensions. As seen in the figure, < T > decreases faster theoretical link between the ratio of static, 0, to high- 2 than Λ . There is little experimental information re- frequency, ∞, permittivity and the ratio of longitudinal, garding the effective mass in this temperature range. A ωL, to transverse, ωT , phonon frequencies. This is the larger temperature-dependent effective mass would make expression first derived by Lyddane, Sachs, and Teller[61, Λ2 smaller and its decrease faster. Thus, one can bridge 62]: the gap between < T > and Λ2 with a hypothetical temperature-dependent polaronic mass. However, there ω 0 = ( L )2 (5) is no experimental evidence for this. At low tempera- ω tures, quantum oscillations[35, 36] resolve a modest mass ∞ T enhancement(≤ 4). This is also the case of infrared When the transverse mode is soft, the system begins conductivity[57]. Thus, we cannot simply picture con- to show a large static polarizability. As one can see in ductivity as ballistic charge transport along Landauer Fig. 6b, the available experimental data confirm the scal- wires. Let us now consider a diffusive picture. ing between these two distinct measurable quantities. A

survive in presence of dilute metallicity. Indeed, neutron 2 1 0 scattering studies have explored the way the soft mode D (this work) responds to the introduction of mobile electrons[65, 66] ω Shirane & Yamada 1 TO and found negligible shift in frequency for carrier densi- ) 3 - 19 −3

V ties below 7.8 10 cm . ) e s / 2 m

4 As one can see in Fig. 6a, comparing our data with m 0 c 1 ( (

those reported by Yamada and Shirane[60], one ﬁnds 3 - D 0 T

that: ω

1 −3 µkBT ∝ ωT 0 ∝ /ωT 0 (6) a) 0 . 1 1 2 0 An explanation for this empirical link between carrier 3 0 b) 1 0 0 mobility and lattice properties is beyond the scope of this paper and emerges as a challenge to theory. 8 0 2 0 C. Concluding remarks ) 2 V

6 0 e ε / m 4 (

The persistence of T-square resistivity in dilute metal- 2 0 0 Τ 1

ω lic strontium titanate is puzzling[22]. However, its con- 4 0 1 0 text is familiar. Electrons are degenerate and kF ` >> 1. We now see that this low-temperature puzzle is only the tip of an iceberg. Above 100 K, the application of the ε (Müller & Burkham) ω (Shirane & Yamada) Drude picture of conductivity yields a paradox: a car- TO 2 0 rier mean-free-path shorter than all conceivable length 1 0 0 1 5 0 2 0 0 2 5 0 3 0 0 scales. Both the magnitude and the functional behavior T(K) of the temperature-induced change in resistivity are very different from what is expected according to the Bloch- Grüneisenlaw and what is seen in silicon and graphene. FIG. 6: Diffusivity, ferroelectric soft mode and per- The aim of this paper is to attract theoretical attention mittivity a) Blue solid circles represent the diffusion con- to a hitherto unnoticed case of strange metallicity. It re- stant extracted from mobility. The dashed line represents a mains to be seen if electron viscosity plays any role in T−1 temperature dependence. Open circles represent the soft −3 diffusive transport. On the experimental side, extending mode frequency, ωTO according to ref.[59]. The two quantities −3 the available data to higher temperatures and lower dop- scale with each other in this temperature range (D ∝ ωTO). −1 2 ing concentration would map the boundaries of metal- b) The inverse of electric permittivity, [38] and ωTO in the −1 2 same temperature range. The scaling is ∝ ωTO. licity and the MIR limit in this system as well as other polar semiconductors. It is a pleasure to thank Sean Hartnoll and Dmitrii theory assuming an anisotropy in the polarizability of Maslov for stimulating discussions. This work has been the oxygen ion[63], the starting point of the polarization supported by an Ile de France regional grant, by Fonds theory of ferroelectricity[64], explains the temperature ESPCI-Paris and by JEIP- Collègede France and funded dependence of ωTO in strontium titanate quantitatively. in part by a QuantEmX grant from ICAM and the The soft mode and the large permittivity are properties Gordon and Betty Moore Foundation through Grant of the insulating parent. However, they are known to GBMF5305 to KB.

[1] N. E. Hussey, K. Takenaka and H. Takagi. Universality [6] S. A. Hartnoll. Theory of universal incoherent metallic of the Mott-Ioffe-Regel limit in metals. Phil. Mag. 84, transport. Nature Physics 11, 54 (2015). 2847 (2004). [7] S. Mukerjee, V. Oganesyan and D. Huse. Statistical the- [2] V. J. Emery and S. A. Kivelson. Superconductivity in ory of transport by strongly interacting lattice fermions. bad metals. Phys. Rev. Lett. 74, 3253(1995) Phys. Rev. B 73, 035113 (2006). [3] S. Sachdev. Bekenstein-Hawking Entropy and Strange [8] N. H. Lindner and A. Auerbach. Conductivity of hard Metals Phys. Rev. X 5, 041025 (2015) core bosons: A paradigm of a bad metal. Phys. Rev. B [4] O. Gunnarsson, M. Calandra and J. E. Han. Colloquium: 81, 054512 (2010). Saturation of electrical resistivity. Rev. Mod. Phys. 75, [9] A.V. Andreev, S. A. Kivelson and B. Spivak. Hydrody- 1085 (2003). namic Description of Transport in Strongly Correlated [5] J. A. N. Bruin, H. Sakai, R. S. Perry and A. P. Mackenzie. Electron Systems. Phys. Rev. Lett. 106, 256804 (2011). Similarity of scattering rates in metals showing T-linear [10] R. A. Davison, K. Schalm and J. Zaanen. Holographic resistivity. Science 339, 804 (2013). duality and the resistivity of strange metals. Phys. Rev. B 89, 245116 (2014). Ratio in Correlated Oxides. J. Phys. Soc. Jpn. 74, 1107 [11] A. Principi and G. Vignale. Violation of the Wiedemann- (2005) Franz law in hydrodynamic electron liquids. Phys Rev [31] M. Swift and C. G. Van de Walle. Conditions for T2 resis- Lett 115, 056603 (2015). tivity from electron-electron scattering. arXiv:1701.04744 [12] N. Pakhira and R. H. McKenzie. Shear viscosity of [32] D. L. Maslov and A. V. Chubukov. Optical response of strongly interacting fermionic quantum fluids. Phys. Rev. correlated electron systems.Rep. Prog. Phys. 80, 026503 B 92, 125103 (2015). (2017). [13] K. Limtragool and P. Phillips. Power-law optical con- [33] J. Son, P. Moetakef, B. Jalan, O. Bierwagen, N. J. ductivity from unparticles: Application to the cuprates. Wright, R. Engel-Herbert and S. Stemmer. Epitaxial Phys. Rev. B 92, 155128 (2015). SrTiO3 films with electron mobilities exceeding 30,000 [14] E. Perepelitsky, A. Galatas, J. Mravlje, R. Zitko,ˆ E. cm2V −1s−1. Nature Materials 9, 482 (2010). Khatami, B. S. Shastry, and A. Georges. Transport and [34] X. Lin, Z. Zhu, B. Fauqué,and K. Behnia. The Fermi optical conductivity in the Hubbard model: A high- surface of the most dilute superconductor. Phys. Rev. X temperature expansion perspective. Phys. Rev. B 94, 3, 021002 (2013). 235115 (2016). [35] S. J. Allen, B. Jalan, S. B. Lee, D. G. Ouellette, G. [15] J.-C. Zhang, E. M. Levenson-Falk, B. J. Ramshaw, D. Khalsa, J. Jaroszynski, S. Stemmer, and A. H. MacDon- A. Bonn, R. Liang, W. N. Hardy, S. A. Hartnoll, A. Ka- ald. Conduction-band edge and Shubnikov-de-Haas effect pitulnik. Anomalous Thermal Diffusivity in Underdoped in low-electron-density SrTiO3. Phys. Rev. B 88, 045114 YBa2Cu3O6+x, arXiv:1610.05845 (2016) (2013). [16] H. P. R. Frederikse W. R. Thurber and W. R. Hosler. [36] X. Lin, G. Bridoux, A. Gourgout, G. Seyfarth, S. Electronic Transpoort in Strontium Titanate. Phys. Rev. Krämer,M. Nardone, B. Fauquéand K. Behnia. Criti- 134, A442 (1964). cal Doping for the Onset of a Two-Band Superconducting [17] O. N. Tufte and P. Chapman. Electron Mobility in Semi- Ground State in SrTiO3−δ. Phys. Rev. Lett. 112, 207002 conducting Strontium Titanate. Phys. Rev. 155, 796 (2014). (1967). [37] K. Behnia, On mobility of electrons in a shallow Fermi [18] H. P. R. Frederikse and W. R. Hosler. Hall Mobility in sea over a rough seafloor. J. Physics: Condens. Matt. 27, SrTiO3. Phys. Rev. 161, 822 (1967). 375501 (2015). [19] S. H. Wemple, M. Didomenico, Jr., and A. Jayara- [38] K. A. Müllerand H. Burkard SrTiO3: An intrinsic quan- man. Electron Scattering in Perovskite-Oxide Ferroelec- tum paraelectric below 4 K. Phys. Rev. B 19, 3593 tric Semiconductors. Phys. Rev. 180, 547 (1969). (1979). [20] A. Spinelli, M. A. Torija, C. Liu, C. Jan and C. Leighton. [39] R. A. Cooper Y. Wang, B. Vignolle, O. J. Lipscombe, Electronic transport in doped SrTiO3: Conduction mech- S. M. Hayden, Y. Tanabe, T. Adachi, Y. Koike, M. No- anisms and potential applications, Phys. Rev. B 81, hara, H. Takagi, C. Proust and N. E. Hussey. Anomalous 155110 (2010). Criticality in the Electrical Resistivity of La2xSrxCuO4. [21] D. van der Marel, J. L. M. van Mechelen and I. I. Mazin. Science 323, 603 (2008). Common Fermi-liquid origin of T2 resistivity and super- [40] Q. Tao, B. Loret, B. Xu, X. Yang, C. W. Rischau, X. conductivity in n-type SrTiO3. Phys. Rev. B 84, 205111 Lin, B. Fauqu, M. J. Verstraete, and K. Behnia. Non- (2011). monotonic anisotropy in charge conduction induced by 2 [22] X. Lin, B. Fauquéand K. Behnia. Scalable T resistivity antiferrodistortive transition in metallic SrTiO3. Phys. in a small single-component Fermi surface. Science 349, Rev. B 94, 035111 (2016). 945 (2015). [41] H. Takagi et al. Systematic Evolution of Temperature- [23] A. Baratoff and G. Binnig. Mechanism of superconduc- Dependent Resistivit in La2−xSrxCu04. Phys. Rev. Lett. tivity in SrTiO3. Physica B+C 108, 1335 (1981). 69, 2975 (1992). [24] A. Verma, A.P. Kajdos, T. A. Cain, S. Stemmer, and [42] N. E. Hussey, K. Nozawa, H. Takagi, S. Adachi D. Jena. Intrinsic Mobility Limiting Mechanisms in and K. Tanabe. Anisotropic resistivity of YBa2Cu4O8: Lanthanum-Doped Strontium Titanate. Phys. Rev. Lett. Incoherent-to-metallic crossover in the out-of-plane 112, 216601 (2014). transport. Phys. Rev. B 56, R11 423(R) (1997). [25] B. Himmetoglu, A. Janotti, H. Peelaers, A. Alkauskas, [43] A. F. Bangura et al. Small Fermi Surface Pockets in Un- and C. G. Van de Walle. First-principles study of the derdoped High Temperature Superconductors: Observa- mobility of SrTiO3. Phys. Rev. B 90, 241204(R) (2014). tion of Shubnikov-de Haas Oscillations in YBa2Cu4O8. [26] E. Mikheev, B. Himmetoglu, A. P. Kajdos, P. Moetakef, Phys. Rev. Lett. 100, 047004 (2008). T. A. Cain, C. G. Van de Walle, and S. Stemmer. Lim- [44] A. W. Tyler, A. P. Mackenzie, S. NishiZaki, and Y. itations to the room temperature mobility of two- and Maeno. High-temperature resistivity of Sr2RuO4: Bad three-dimensional electron liquids in SrTiO3. Appl. Phys. metallic transport in a good metal. Phys. Rev. B 58, Lett. 106, 062102 (2015) R10107(R) (1998) [27] W. X. Zhou et al. Electron–soft phonon scattering in n [45] D. K. Efetov and P. Kim. Controlling Electron-Phonon -type SrTiO3. Phys. Rev. B 94, 195122 (2016). Interactions in Graphene at Ultrahigh Carrier Densities. [28] H. K. Pal, V. I. Yudson, and D. L. Maslov, Lith. J. Phys. Phys. Rev. Lett. 105, 256805 ( 2010). 52, 142 (2012). [46] C. Yamanouchi, K. Mizuguchi and W. Sasaki. Electric [29] K. Kadowaki, S.B. Woods. Universal relationship of the Conduction in Phosphorus Doped Silicon at Low Tem- resistivity and specific heat in heavy-fermion compounds. peratures. J. Phys. Soc. Jpn. 22, 859 (1967). Solid State Commun. 58, 507 (1986). [47] S. H. Wemple. Some Transport Properties of Oxygen- [30] N. E. Hussey. Non-generality of the Kadowaki–Woods Deficient Single-Crystal Potassium Tantalate (KTaO3). Phys. Rev. 137, A1575 (1965). [48] R. L. Petritz and W. W. Scanlon. Mobility of Electrons Bosman and R. J. J. Zijlstra. Diffusion coefficient of elec- and Holes in the Polar Crystal, PbS. Phys. Rev. 97, 1620 trons in silicon. J. Appl. Phys. 52, 6713 (1981) (1955) [59] P. A. Fleury and J. M. Worlock. Electric-Field-Induced [49] R. S. Allgaier and W. W. Scanlon. Mobility of Electrons Raman Scattering in SrTiO3 and KTaO3. Phys. Rev. and Holes in PbS, PbSe, and PbTe between Room Tem- 174, 613 (1968) perature and 4.2K. Phys. Rev. 111, 1029 (1958) [60] Y. Yamada, and G. Shirane. Neutron Scattering and Na- [50] N. F. Mott and R. W. Gurney. Electronic Processes in ture of the Soft Optical Phonon in SrTiO3. J. Phys. Soc. Ionic Crystals. Clarendon Press, Oxford (1940) Jpn. 26, 396(1969) [51] P. P. Edwards and M. J. Sienko. Universality aspects of [61] R. H. Lyddane, R. G. Sachs, and E. Teller. On the Polar the metal-nonmetal transition in condensed media. Phys. Vibrations of Alkali Halides. Phys. Rev. 59, 673 (1941) Rev. B 17, 2575 (1978). [62] W. Cochran and R.A. Cowley. Dielectric constants and [52] N. F. Mott. Metal-insulator transitions. Second edition, lattice vibrations. J. Phys. and Chem. Sol. 23, 447 (1962) Taylor and Francis, London (1990) [63] R. Migoni, H. Bilz, and D. D. Bäuerle.Origin of Raman [53] R. P. Lowndes and A. Rastogi. Stabilization of the para- Scattering and Ferroelectricity in Oxidic Perovskites. electric phase of KTaO3 and SrTiO3 by strong quartic an- Phys. Rev. Lett. 37, 1155 (1976) harmonicity. J. Phys. C: Solid State Phys. 6, 932 (1973). [64] H. Bilz, G. Benedek, and A. Bussmann-Holder. Theory [54] N. W. Ashcroft and N. D. Mermin, Solid State Physics, of ferroelectricity: The polarizability model. Phys. Rev. New York : Holt, Rinehart and Winston (1976). B 35, 4840 (1987) [55] Y. Imry Y. and R. Landauer. Conductance viewed as [65] D. Bäuerle,D. Wagner, M. Wöhlecke, B. Dorner and H. transmission. Rev. Modern Phys. 71, S306 (1999). Kraxenberger. Soft modes in semiconducting SrTiO3: II. [56] D. M. Eagles. Polaron coupling constants in SrTiO3. J. The ferroelectric mode. Z. Physik B - Condensed Matter Phys. Chem. Solids 26, 672 (1965) 38, 335 (1980) [57] J.L.M van Mechelen, D. van der Marel, C. Grimaldi, A. [66] A. Bussmann-Holder, H. Bilz, D. Bäuerle,D. Wagner. A B. Kuzmenko, N. P. Armitage, N. Reyren, H. Hagemann, polarizability model for the ferroelectric mode in semi- and I. I. Mazin. Electron-phonon interaction and charge conducting SrTiO3. Z. Physik B - Condensed Matter 41, carrier mass enhancement in SrTiO3. Phys. Rev. Lett. 353 (1981) 100, 226403 (2008). [58] R. Brunetti, C. Jacoboni, F. Nava, L. Reggiani, G.