Papers in Physics, vol. 7, art.

070009 (2015)

www.papersinphysics.org

Received: 17 May 2015, Accepted: 12 June 2015

Edited by: A. Vindigni
Reviewed by: M. Perfetti, Dipartimento di Chimica, Universitá di Firenze, Italy
Licence: Creative Commons Attribution 3.0
DOI: http://dx.doi.org/10.4279/PIP.070009
ISSN 1852-4249

An intermediate state between the kagome-ice and the fully polarized

state in Dy2 Ti2 O7

S. A. Grigera,1, 2∗ R. A. Borzi,3 D. G. Slobinsky,1† A. S. Gibbs,1

R. Higashinaka,4 Y. Maeno,5 T. S. Grigera3

Dy2 Ti2 O7 is at present the cleanest example of a spin-ice material. Previous theoretical
and experimental work on the first-order transition between the kagome-ice and the fully
polarized state has been taken as a validation for the dipolar spin-ice model. Here we inves-
tigate in further depth this phase transition using ac-susceptibility and dc-magnetization,
and compare this results with Monte-Carlo simulations and previous magnetization and
specific heat measurements. We find signatures of an intermediate state between the
kagome-ice and full polarization. This signatures are absent in current theoretical models
used to describe spin-ice materials.

I. Introduction angular lattices along the [111] direction. The spins

sit at the vertices of tetrahedra and can point ei-
Spin-ice materials are deceptively simple in their ther to their center or towards the outside. The
constitution: classical Ising spins with nearest- magnetic frustration can be seen at the level of a
neighbour ferromagnetic interactions forming a py- single tetrahedron: the energy is minimized by hav-
rochlore lattice. This crystal structure can be ing two spins pointing inwards and two outwards.
thought as an alternating stack of kagome and tri- This is the ice rule, which corresponds exactly to
∗ E-mail:
the Pauling rules for protons in water ice; like the
[email protected]
† Now
at: Departamento de Ingenierı́a Mecánica, Facul-
latter, it also leads to zero-point entropy, a charac-
tad Regional La Plata, Universidad Tecnológica Nacional, teristic signature of spin-ice systems [1].
1900 La Plata, Argentina. We have chosen to work on Dy2 Ti2 O7 as the
cleanest example of a spin-ice material. Its ground
1 School of Physics and Astronomy, University of St An- state properties can be well described by a model
drews, North Haugh, St Andrews KY16 9SS, UK
with only an effective nearest neighbour exchange
2 Instituto de Fı́sica de Lı́quidos y Sistemas Biológicos, interaction Jsi of ≈ 1.1 K [2]. Within this frame-
UNLP-CONICET, 1900 La Plata, Argentina
work, when one applies an external magnetic field
3 Instituto de Investigaciones Fisicoquı́micas Teóricas H in [111] below 1 K, the polarization of the sys-
y Aplicadas UNLP-CONICET and Departamento de
Fı́sica, Facultad de Ciencias Exactas, Universidad Na-
tem will happen in two steps. First, the spins in the
cional de La Plata, 1900 La Plata, Argentina triangular lattice that lie parallel to [111] will ori-
4 Graduate School of Science, Tokyo Metropolitan Univer-
ent along the magnetic field, removing part of the
sity, Hachioji, Tokyo 192-0397, Japan. residual entropy but with no change in the configu-
5 Department of Physics, Kyoto University, Kyoto 606-
rational energy [3, 4]. When the magnetic moment
8502, Japan. of this sublattice has saturated, the magnetization
M cannot be further increased without breaking

070009-1
 Papers in Physics, vol. 7, art. 070009 (2015) / S. A. Grigera et al.

the spin-ice rule. This leads to a plateau as a func- other words, almost full order is achieved in the sys-
tion of field at M = 3.33 µB /Dy-ion, characteris- tem for temperatures just below Tc and a magnetic
tic of the kagome ice state. At higher fields, the field of 1 T.
spins in the kagome lattice are finally fully polar- Specific heat Cp measurements confirmed the ex-
ized, leading to a sudden but continuous increase istence of a critical end-point —a sharp peak is
in M towards its saturation. This behavior was clearly seen very near the precise spot in field and
predicted theoretically and found in Monte Carlo temperature specified by Sakakibara et al. [11].
simulations [5, 6]. In spite of this, something differ- However, the identification of a single first-order
ent happens in real spin-ice materials. line below Tc is less clear. The Cp (T )vs.H curves
Since the magnetic moment of the magnetic ions show peaks at the fields Hc (T ) identified in [9] as
in spin-ice materials is quite large – the one asso- the first-order line, albeit of much smaller ampli-
ciated with Dy3+ ions in Dy2 Ti2 O7 is near 10µB – tude than that at Tc . Additionally, below 300 mK,
, long range dipolar interactions have to be con- a second peak at higher fields is discernible [11].
sidered [7]. These interactions do not alter the Even at the lowest temperatures (100 mK), mag-
zero field ground state [8], but have a big effect netic fields above 2 T are needed to coerce the spe-
on its excitations. In relation to this, the transi- cific heat down to 0. This suggests that, in spite
tion to the fully polarized state —which is the main of the absence of thermal excitations, the system
concern of this paper— experiments a qualitative does not reach full polarization immediately after
change. T. Sakakibara and collaborators [9] studied the first order transition from the kagome ice, and
experimentally the magnetization with H// [111] an intermediate state establishes between these two
to temperatures much smaller than Jsi . After a well-known phases. This specific heat features were
well defined plateau at ≈ 3.33 µB /Dy-ion, they ob- confirmed by ac-susceptibility measurements on the
served a very sharp increase in the magnetization. same samples [12]. In all cases, the sample sat at a
The presence of hysteresis was a convincing argu- fixed platform with respect to the magnetic field
ment that the real system reaches the fully polar- and therefore the alignment with respect to the
ized state through a metamagnetic first order phase [111] direction was within a few degrees. An an-
change at the lowest temperatures. The change gular dependent study of the magnetization with
in M becomes continuous at the critical end-point Sato and coworkers [19] showed these asymmetries,
Tc = 360 ± 20mK and µ0 Hc ≈ 0.93 T [9]. and additional features in the polarization tran-
The change in character of this transition — sition were seen at small angles away from [111].
from a crossover to a discontinuity when dipolar The implications of these results in the current un-
interactions are included— was later understood in derstanding and modeling of the spin-ice materials
terms of the defects associated with the breaking of have not been considered.
the ice rules, or monopoles. The nearest neighbors In this paper, we study in detail this additional
model corresponds to the case of free non-conserved intermediate state, and show that it cannot be ex-
monopoles sitting in a diamond lattice. Including plained by any of the models currently used to
dipolar interactions implies turning on a Coulomb study spin-ice materials. Working at small angles
interaction between these charges, allowing them away from [111], we looked for a magnetic signa-
to condense through a real first order transition ture by repeating the static magnetization mea-
[10]. Numerical simulations (including Ewald sum- surements in several samples. In addition, improv-
mations to take into account long range interac- ing the sensitivity by three orders of magnitude, we
tions) proved this picture right, and provided an measured ac-susceptibility at different frequencies,
additional validation to the dipolar model [10]. The which also allowed us to do a characterisation of
M vs. H curves obtained in these simulations are the dynamics of the observed transitions. In order
quite symmetrical around Hc . The jump in the to gain further insight into this possible interme-
magnetization ∆M (T ) when crossing the first order diate state, we performed Monte Carlo simulations
transition line grows very abruptly with decreasing of the experimental situation using the currently
temperature: for T only ≈ 10% below Tc , ∆M (T ) accepted models including Ewald summations and
amounts to ≈ 90% of the total change in magne- exchange interactions up to the third nearest neigh-
tization from the kagome ice to full saturation. In bor [13, 14].

070009-2
 Papers in Physics, vol. 7, art. 070009 (2015) / S. A. Grigera et al.

Figure 1: Real (left) and Imaginary (right) parts of the ac-susceptibility as a function of temperature and
magnetic field, for temperatures between 50 mK and 6000 mK and magnetic fields between -4 and 4 T.
The oscillatory field was of amplitude 0.05 Oe and at a frequency of 87Hz. The zero-field Schottky-type
anomaly corresponding to the onset of spin-ice correlations, and the peaks corresponding to the critical
point at ±1 T and ≈400 mK are clearly seen.

II. Methods whole area of interest. The excitation frequency

in this case is ω = 87 Hz; the main features we
For our work, we measured several Dy2 Ti2 O7 single describe in the following are qualitatively indepen-
crystals grown in Kyoto and in St Andrews with the dent of ω. At zero field, there is a very noticeable
floating-zone method. Samples were oriented us- peak in both ∆χ0 and ∆χ00 for T ≈ 2 K. This cor-
ing Laue diffraction and cut into 3 mm long prisms responds to the Schottky-type anomaly associated
of square or octagonal section of approximately with the onset of spin-ice correlations of the system.
1 mm2 , with the [111] direction along the long axis The magnetic field axis spans from -4 to 4 T, and
to reduce demagnetising effects with the field in we can clearly see in the real part two peaks (at
the vicinity of [111] (≈ 5o ). The experiments were positive and negative fields) corresponding to the
performed in a dilution refrigerator in St Andrews. critical point at ≈ ±1 T and ≈ 400 mK. For tem-
Samples were thermally grounded to the mixing peratures below 400 mK, we see a much smaller fea-
chamber through gold wires attached into them ture in ∆χ00 , which has a correspondence in ∆χ00 :
with silver paint. For susceptibility, we used a a ridge with an amplitude that decreases as a func-
drive field of 3.3 · 10−5 T r.m.s., and counter-wound tion of temperature. The magnitude of the latter is
pickup coils each consisting of approximately 1000 comparatively very small. At low temperatures and
turns of 12 µm diameter copper wire. The filling for fields 0.3 T< |µ0 H| < 0.9 T, and 2 T < |µ0 H|,
factor of the sample in the pick up coil was of ap- the susceptibility is very low, in accordance to the
proximately 90%. We measured using drive fields of kagome ice plateau and the saturation in the mag-
frequencies varying from approx. 10 Hz to 1.0 kHz. netization, respectively.
Low temperature transformers mounted on the 1 K
pot of the dilution refrigerator were used through- We now concentrate on the real part of the sus-
out to provide an initial signal boost of approx- ceptibility at temperatures below Jsi . In Fig. 2,
imately a factor of 100. The magnetization was we can see a series of curves at fixed temperatures
measured using a home-built capacitance Faraday (from 50 to 500 mK) and fields between −3.5 and
magnetometer [15]. 3.5 T. The excitation field used was 0.05 Oe, and
the frequency 87 Hz. The curves have been off-
set by 30% for clarity. The field was swept from
III. Results and Discussion negative to positive values. Before the kagome
ice state is established, the low field susceptibility
Figure 1 shows the real (∆χ0 , left) and imaginary (|µ0 H| < 0.3 T) at temperatures below 600 mK is
(∆χ00 , right) parts of the ac-susceptibility χ as a strongly dependent on the magnetic field sweep rate
function of temperature and magnetic field in the and direction (increasing or decreasing), both signs

070009-3
 Papers in Physics, vol. 7, art. 070009 (2015) / S. A. Grigera et al.

Figure 3: Phase diagram with field slightly tilted

Figure 2: Low temperature real part of the ac- from [111] (θ ≤ 5o ). An intermediate phase is seen
susceptibility as a function of field at fixed temper- between the kagome-ice and fully polarized regions.
atures as indicated in the plot. The excitation field The black circle is the critical point as identified
was 0.05 Oe at a frequency of 87Hz. The curves are from a peak in the real part of the ac-susceptibility,
offset by 30% for clarity. As temperature is lowered χ0 . The dotted white circles correspond to a small
from 400 mK, the peak at approx. 1 T at 400mK doubled peaks seen in χ0 with a corresponding fea-
rapidly decreases in amplitude, and splits into two ture in the imaginary part χ00 , while the white cir-
peaks at lower temperatures. cles denote small peak in χ0 with no signature in χ00 .
The red circles are taken from peaks in the specific
heat (C) measurements of reference [11]. The main
of out-of-equilibrium behavior. At higher magnetic divergence of C seen in reference [11] and identified
fields, we only observe a small difference in the as a critical point coincides with the critical point
height of the peaks at around ±1 T, depending on (black circle).
whether the transitions are swept upwards or down-
wards in field. The position changes very little,
and the shape of the features is unaltered. As we The coincidence between these two experiments of
lower the temperature, the peak at ≈ 1 T decreases different quantities, on different samples, laborato-
markedly in amplitude, but without a correspond- ries and experimental setup is remarkable.
ing change in its high field side shoulder. Below As mentioned before, this secondary peak at
400 mK, it eventually splits into two distinct fea- higher fields is absent in the dM/dH data presented
tures. Their separation in the field axis (≈ 0.1 T on Ref. [9]. We measured the magnetization using
at 300 mK) is consistent with previous measure- a Faraday balance on the same samples and un-
ments for a similar sample orientation with respect der similar temperature and field conditions than
to [111] [19]. While the first set of peaks has a before [16]. The main body of Fig. [17] shows our
correlate in the imaginary part of ∆χ (not shown dM/dH as a function of field, compared with curves
here), no feature is discernible in ∆χ00 for the peaks of ∆χ at T = 100 mK and frequencies spanning two
at higher fields. orders of magnitude (from ω ≈ 10 to 1000 Hz). For
In Fig. 3, we have plotted the position of these clarity, we have multiplied ∆χ by a factor of twenty.
peaks as a function of field and temperature (white The peak in dM/dH is markedly asymmetric, with
circles), and the position of the critical point (black an extended tail in the high field side but no addi-
circle). We have taken the specific heat data from tional feature is seen at high fields, in coincidence
reference [11] and determined the position of the with Sakakibara’s observations. On the other hand,
peaks in C vs. H for different temperatures. These the second peak is clearly seen for low tempera-
are plotted in this same graphic as red symbols. ture (T < 400 mK) at all measured frequencies in

070009-4
 Papers in Physics, vol. 7, art. 070009 (2015) / S. A. Grigera et al.

25
in [111] was of a single transition—the “dimer to
dM/dH
monomer” transition of refs. [6, 18]. A small ad-

/ T Dy]
20

20 ditional perpendicular field – present in the exper-

/ T Dy]

B
10
iments at small angles away from [111]– induces
[ order in the dimers in the kagome-ice state, but
15

0
does not change the prediction of a single transi-
B

0.5 1.0 1.5

tion into the fully polarised “monomer” state [18].
[

10
This might hold true when further interactions are
added, such as dipolar or further neighbor exchange
X 20
5
interactions. In order to investigate this, we per-
formed a numerical check. We did extensive Monte
0 Carlo simulations of the dipolar model including
0.5 1.0 1.5 2.0
Ewald summations to account for the dipolar long
Field (tesla) range interactions. We also added exchange inter-
actions up to third nearest neighbors (taking the
Figure 4: dM/dH (dotted line) and real part of ac- exchange constants and other parameters within
susceptibility measured at different excitation fre- the constraints given by refs. [13, 14]). We ex-
quencies, from top to bottom: 19, 37, 77, 136, 277, plored a wide range of field angles around [111], but
561, and 1117 Hz, and for T = 100mK. For ease were unable to detect a double feature in CV at low
of comparison, the latter have been multiplied by a temperatures compatible with the experimental ob-
factor 20, and normalized to the amplitude of sec- servations. It is then worth stressing that the very
ond peak (no imaginary part has been measured observation of a second feature –even when taking
for this feature). The inset shows both sets of data into account a possible sample misalignment– asks
in the same scale. for new ingredients in the Hamiltonians that are
regularly used to describe spin-ice materials.

the ac-susceptibility. While these two experiments

seem to be in mutual contradiction, the issue can be Given these considerations, it is difficult to dis-
easily explained in terms of the resolutions of both cuss on the nature of this intermediate state. It is
techniques. The inset of Fig. 4 shows both sets of tempting to think of some sort of “charge” ordering
data on the same scale; we can see that the high in the diamond lattice (2-in 2-out tetrahedra within
field shoulder on the dM/dH peak directly corre- a majority of 3-1 and 1-3), previous to the final
sponds (in the limit of long measurement times or Zn-blende arrangement, where only ≈ 40 − 50% of
low frequencies) to the second peak detected with the sites are occupied by single monopoles. Note
ac-susceptibility. that this does not rule out the possibility of still
Through this analysis, we can see that the exper- storing some residual entropy, since there are dif-
imental volume of data concerning this transition ferent spin configurations that generate the same
seems to be compatible. Between ≈ 300mK and charge within a given tetrahedron. We have not
the lowest temperatures (50 mK in Ref. [9]), only found previous data of the evolution of the entropy
≈ 60% of the total change in magnetization occurs as a function of field at temperatures well below
when traversing the first order transition line. The Tc . However, the very asymmetric shape of the en-
remaining 40% is delivered gradually when the field tropy at 350 mK obtained using the magnetocaloric
is further increased to values well above 1.5Hc , in effect shows that at this temperature the system is
a fashion that does not seem to depend much on already experiencing a strong first order metamag-
temperature (see Fig. 3 on Ref. [9]). This gradual netic transition, as mentioned in Ref. [20]. This
(as opposed to discontinuous) change is behind the work shows that a big fraction of the residual en-
asymmetric shape of the magnetization curves, and tropy stored in the kagome planes remains in the
the second set of peaks in Cp and ∆χ. system well above Hc [20], suggesting that the in-
The theoretical prediction for the transition be- termediate state is indeed a partially disordered
tween kagome-ice to fully polarized state with field one.

070009-5
 Papers in Physics, vol. 7, art. 070009 (2015) / S. A. Grigera et al.

IV. Conclusions [111] magnetic field, Phys. Rev. B 70, 104418

(2004).
In conclusion, we observe an intermediate state be-
tween the kagome-ice and the fully polarized state [7] S T Bramwell, M J Harris, B C den Hertog,
when the field is slightly tilted from the [111] di- M J P Gingras, J S Gardner, D F McMorrow,
rection. The signature of a double step we find A R Wildes, A L Cornelius, J D M Cham-
in ac-susceptibility and magnetization measure- pion, R G Melko, T Fennell, Spin correlations
ments is also present in earlier calorimetric mea- in Ho2 Ti2 O7 : A dipolar spin ice system Phys.
surements, and suggested by magnetocaloric effect Rev. Lett. 87, 047205 (2001).
experiments. This feature cannot be captured by
the models regularly used to describe spin-ice sys- [8] S V Isakov, R Moessner, S L Sondhi, Why
tems, fact that asks for further model refinements. spin ice obeys the ice rules Phys. Rev. Lett. 95,
At present, this data stands as a challenge for the 217201 (2005).
development of a realistic theoretical model of spin-
[9] T Sakakibara, T Tayama, Z Hiroi, K Mat-
ice materials.
suhira, S Takagi, Observation of a liquid-gas-type
transition in the Pyrochlore spin ice compound
Dy2 Ti2 O7 in a magnetic field, Phys. Rev. Lett.
90, 207205 (2003).
Acknowledgements - We thank Joseph Betouras,
Andrew Green and Chris Hooley for useful discus- [10] C Castelnovo, R Moessner, S L Sondhi, Mag-
sions. SAG would like to acknowledge financial netic monopoles in spin ice, Nature 451, 42
support from the Royal Society (UK), RAB and (2008).
TSG from CONICET, UNLP and ANPCYT (Ar-
gentina). [11] R Higashinaka, H Fukazawa, K Deguchi,
Y Maeno, Low temperature specific heat of
Dy2 Ti2 O7 in the kagome ice state, J. Phys. Soc.
[1] M J Harris, S T Bramwell, D F McMorrow, T Jpn. 73, 2845 (2004).
Zeiske, K W Godfrey, Geometrical frustration in
the ferromagnetic Pyrochlore Ho2 Ti2 O7 , Phys. [12] R Higashinaka, Field orientation control of ge-
Rev. Lett. 79, 2554 (1997). ometrical frustration in the spin ice Dy2 Ti2 O7 ,
Ph. D. thesis, Kyoto University (2005).
[2] S T Bramwell, M J P Gingras, Spin ice state
in frustrated magnetic Pyrochlore materials, Sci- [13] J P C Ruff, R G Melko, M J P Gingras, Finite-
ence 294, 1495 (2001). temperature transitions in dipolar spin ice in a
[3] Z Hiroi, K Matsuhira, S Takagi, T Tayama, T large magnetic field, Phys. Rev. Lett. 95, 097202
Sakakibara, Specific heat of kagom ice in the Py- (2005).
rochlore Oxide Dy2 Ti2 O7 , J. Phys. Soc. Jpn. 72, [14] T Yavors’kii, T Fennell, M J P Gingras, S T
411 (2003). Bramwell, Dy2 Ti2 O7 spin ice: A test case for
[4] M Udagawa, M Ogata, Z Hiroi, Exact result of emergent clusters in a frustrated magnet, Phys.
ground-state entropy for Ising Pyrochlore mag- Rev. Lett. 101, 037204 (2008).
nets under a magnetic field along [111] axis, J.
[15] D Slobinsky, R A Borzi, A P Mackenzie, S
Phys. Soc. Jpn. 71, 2365 (2002).
A Grigera, Fast sweep-rate plastic Faraday force
[5] M J Harris, S T Bramwell, P C W. Holdsworth, magnetometer with simultaneous sample tem-
J D M Champion, Liquid-gas critical behavior in perature measurement , Rev. Sci. Instrum. 83,
a frustrated Pyrochlore ferromagnet, Phys. Rev. 125104 (2012).
Lett. 81, 4496 (1998).
[16] Since we are using two different probes, the ex-
[6] S V Isakov, K S Raman, R Moessner, S L act position and orientation of the sample with
Sondhi, Magnetization curve of spin ice in a respect to the magnet differs slightly between the

070009-6
 Papers in Physics, vol. 7, art. 070009 (2015) / S. A. Grigera et al.

ac-susceptibility and the magnetization measure- lowering the temperature, the area loss observed
ment, and a positive shift of 0.02 T in the magne- in this figures is quite natural.
tization measurement was necessary to make the
critical field of the first order transition coincide. [18] R Moessner, S L Sondhi, Theory of the [111]
magnetization plateau in spin ice, Phys. Rev. B
[17] The jump in magnetization ∆M is essentially 68, 064411 (2003).
independent of temperature at low T . Since ∆M
is the integral of the susceptibility, one would [19] H Sato, K Matsuhira, T Sakakibara, T
naively expect that area below the peak in χ0 Tayama, Z Hiroi, S Takagi, Field-angle depen-
(Fig. 2) to be also independent of T . But this dence of the ice-rule breaking spin-flip transi-
is true only for the dc susceptibility, or, more tion in Dy2 Ti2 O7 , J. Phys. Condens. Matter 19,
0
accurately, for χ measured at frequencies lower 145272 (2007).
than the inverse of the longest relaxation time. [20] H Aoki, T Sakakibara, K Matsuhira, and Z Hi-
The fact that we can measure an out of phase roi, Magnetocaloric effect study on the Pyrochlore
response ∆χ00 reveals we are actually measuring spin ice compound Dy2 Ti2 O7 in a [111] magnetic
dynamic response, i.e. that our frequencies are field, J. Phys. Soc. Jpn. 73, 2851 (2004).
high and some relaxation processes do not con-
tribute to ∆χ0 . Since relaxation times grow on

070009-7