Room-Temperature Ferromagnetism of

Graphene
Yan Wang,

Yi Huang,*
,
You Song,

Xiaoyan Zhang,

Yanfeng Ma,

Jiajie Liang,

and Yongsheng Chen*

,
Key Laboratory of Functional Polymer Materials and Center for Nanoscale Science
and Technology, Institute of Polymer Chemistry, College of Chemistry, Nankai
UniVersity, Tianjin 300071, China, and State Key Laboratory of Coordination
Chemistry, School of Chemistry and Chemical Engineering, Nanjing UniVersity,
Nanjing 210093, China
Received September 16, 2008; Revised Manuscript Received November 25, 2008
ABSTRACT
Aiming at molecular-based magnets, ferromagnetism of pure carbon-based materials is fundamentally and technologically extremely important
for many applications. While it is still not fully understood, many recent theoretical works have suggested that one-atom-thick two-dimensional
graphene materials may show ferromagnetism due to the existence of various defects or topological structures as the spin units and the
possible long-range ordered coupling among them. Here, we report the experimental results on the ferromagnetism of graphene-based materials
at room temperature. The observed room-temperature ferromagnetism is believed to come from the defects on graphene.
Organic magnets are both fundamentally and technologically
extremely important and have been sought for many years.
1
However, ferromagnetism in organic materials is usually
observed only at very low temperatures,
2,3
and the reports
about the existence of intrinsic magnetism of pure carbon-
based materials,
4,5
which have only s/p electrons, have been
much controversial due to trace amounts of Fe or other
magnetic impurities. The scientic community became much
more aware of the possible existence of room-temperature
carbon-based and metal-free organic ferromagnets since the
serendipitous discovery of strong signals in rhombohedral
C
60
was reported in 2001.
6
While it was retracted in 2006,
the article did raise (again) the possibility and expectation
of pure carbon-based materials with Curie temperature above
room temperature. Several groups have reported the experi-
mental ferromagnetism of highly oriented pyrolytic graphite
(HOPG) materials.
4,5,7
A spongy, ultralightweight form of
carbon with a strong but temporary ferromagnetism was also
reported in 2004.
8
Esquinzai et al. have reported that HOPG
samples bombarded with protons appeared ferro- or ferri-
magnetism with a stable magnetic ordering at room temper-
ature.
7
Their results have provided a possible answer to the
origin of ferromagnetism in other carbon-based structures,
which is not correlated with the magnetic impurities. This
kind of magnetic behavior may be ascribed to itinerant
electron magnetism rather than superexchange magnetic
coupling interaction between spins like molecule-based
magnets. Many theoretical studies
9,10
have also suggested the
highly possible existence of ferromagnetism for pure carbon
materials, even at room temperature. The recent discovery
of graphene,
11
one-atom-thick two-dimensional (2D) layers
of sp
2
-bonded carbon, and its many unusual properties have
prompted intensive studies for this 2D nanoscale pure carbon
material. In addition to their extraordinary mechanical
12
and
electronic transport
13,14
properties, numerous reports have
predicted ferromagnetic ordering could exist among various
defects on graphene structures, such as vacancy, topological
defects or frustration, and hydrogen chemisorption or perhaps
the edge.
4,15-19
Interestingly, the proposed simplest building
block
17,18
for these graphene fragments with nonzero net spin
has a phenalenyl molecule structure unit and quite some of
these stable radical organic molecules with this or similar
structures have been synthesized and characterized elegantly
by Haddon and other groups.
20-22
At solid state, these pure
organic magnetic materials also show interesting semicon-
ducting properties. The large -conjugation in these radical
organic molecules offers the required stability for the
unpaired electrons in these structures as net spin units.
20-22
Also, the molecular structure of bulk graphene materials
produced using wet chemistry
23-25
from graphite is similar
(but much larger) with many large fused polyaromatic
graphene molecules reported by Mullen and other
* Corresponding authors, yschen99@nankai.edu.cn and yihuang@
nankai,edu.cn.

Key Laboratory of Functional Polymer Materials and Center for

Nanoscale Science and Technology, Institute of Polymer Chemistry, College
of Chemistry, Nankai University.

State Key Laboratory of Coordination Chemistry, School of Chemistry

and Chemical Engineering, Nanjing University.
NANO
LETTERS
2009
Vol. 9, No. 1
220-224
10.1021/nl802810g CCC: $40.75 2009 American Chemical Society
Published on Web 12/12/2008
groups.
26
And these materials, with 2D graphene molecular
structures, are rather stable at ambient condition with many
interesting electronic and optoelectronic properties. Further-
more, it has been shown theoretically that zigzag-edged
graphene akes with topological structures can have ferro-
magnetic order and linearly scaling net spin and these
structures are stable at room temperature with an energy gap
of 0.3-0.5 eV at their Fermi level.
17
Very recently, we have
reported the results of transparent electrodes
13
and photo-
voltaic devices
27
using these bulk solution-processable func-
tionalized graphene (SPFGraphene) materials. In this Letter,
we report the room-temperature ferromagnetism of graphene
based on these soluble graphene materials.
Perfect graphene does not exist naturally, but bulk and
solution-processable functionalized graphene materials in-
cluding graphene oxide (GO) can now be prepared.
13,23-25
The graphene material used in this study was prepared from
GO, which was prepared by the modied Hummers method
28
from graphite as described elsewhere.
27
GO can be dispersed
and processed at individual sheet level in water.
27,29
The
graphene samples for magnetic studies in this work were
prepared by two steps. GO was rst reduced using hydrazine
to get the partially reduced graphene oxide sample, and then
it was annealed at different temperatures under an argon
environment to generate the nal products for the magnetiza-
tion measurement. Three samples, Graphene-400, Graphene-
600, and Graphene-800, were prepared at annealing tem-
perature of 400, 600, and 800 C, respectively. Details of
the experimental procedure are described in Methods below.
The magnetization measurements were performed with a
SQUID XL-7 magnetometer from Quantum Design with the
reciprocating sample option and a sensitivity of e10
-7
emu.
Figure 1 shows the magnetization hysteresis loops mea-
sured at temperature of 300 K in the eld range of -10 kOe
< H < +10 kOe for the two samples of Graphene-400 and
Graphene-600, where saturation of the magnetization is
clearly seen at about M
s
) 0.004 and 0.020 emu/g after
subtracting the diamagnetic background. The magnetization
was saturated at 1500 and 6000 Oe for samples Graphene-
400 and Graphene-600, respectively. In a smaller eld range
as shown in Figure 2, an enlarged and clearer magnetization
hysteresis loop was observed for both samples of Graphene-
400 and Graphene-600.
These results clearly show that the graphene samples have
a weak ferromagnetic ordering at room temperature. So we
further measured the magnetization of the same samples at
2 K (Figures 3 and 4). As Figure 3 shows, the saturation
magnetization M
s
reaches 0.25 and 0.90 emu/g for Graphene-
400 and Graphene-600, respectively, signicantly increased
compared with that at room temperature. We have also
measured the sample Graphene-800 prepared at 800 C from
the same precursor sample as for Graphene-400 and Graphene-
600. However, the Graphene-800 sample did not show clear
ferromagnetism. The reason for this is yet to be understood.
The reports about the existence of intrinsic magnetism of
pure organic materials have been very controversial, and the
purity of the material is critical. We took great care to
eliminate any possible ferromagnetic impurities. Since all
the three samples (Graphene-400, Graphene-600, and
Graphene-800) came from the same source, e.g., the partially
reduced graphene oxide, the magnetic contribution from the
possible metallic impurity should be similar for all these three
samples. So the negative result for ferromagnetism of
Graphene-800 suggests that metallic impurity should not be
the reason (or at least the major one) of the observed
ferromagnetism of Graphene-400 and Graphene-600. The
expected M
s
value would be 502 emu/g,
30
assuming all the
carbon atoms have a ferromagnetic unpaired electron. Thus,
the observed M
s
(0.9 emu/g) value at 2 K for sample
Graphene-600 corresponds to 0.2% carbon atoms.
To further check the contribution from magnetic impuri-
ties, magnetic measurements were also carried out for the
GO samples and those after only hydrazine reduction or
annealing and no ferromagnetism was observed for these
three samples. The total amount of magnetic (Fe, Ni, Co)
impurities by atomic absorption spectrometry (AAS) analysis
was 48.6 ppm for the partially reduced graphene oxide
samples before annealing. More importantly, the total amount
of magnetic impurities by AAS was 86 ppm for Graphene-
600 after annealing. The increase of magnetic impurity may
be ascribed to the weight loss for the functional groups. The
observed M
s
(0.9 emu/g) value for sample Graphene-600 at
2 K is equivalent to 4700 ppm Fe, 2 orders greater than
the AAS analysis result. Furthermore, we have carried out
the magnetic measurements for ve batches of samples on
three different SQUID instruments, and the results are quite
repeatable. All these results indicate it is very likely that the
room-temperature ferromagnetism comes from the intrinsic
properties of graphene-based materials.
As mentioned above, many theoretical works have been
published about the possible ferromagnetism and its origin
for graphene. But while all the theoretical works so far show
that it is quite easy to nucleate magnetic moments in the
defective sp
2
carbon samples, the fact that they order
ferromagnetically is yet to be fully understood. The structure
of GO contains many C-O-C (epoxide), C-OH, -COOH,
and CdO groups on the main graphene skeleton.
13,29
Initial
reduction of GO with hydrazine would remove some of these
groups and restore some of the damaged sp
2
carbon conjuga-
tion during the GO preparation process and, thus, increase
its electrical conductivity.
31
During further annealing process
Figure 1. Magnetization hysteresis loops at 300 K in the range of
-10 kOe < H < +10 kOe. (a) Sample Graphene-400 (0), M
s
)
0.004 emu/g. (b) Sample Graphene-600 (b), M
s
) 0.020 emu/g.
Nano Lett., Vol. 9, No. 1, 2009 221
under argon, most if not all of the functional groups would
be removed from the graphene sheets. Thus, the isolated
-conjugation area on the pristine graphene sheet should be
restored (or partially) again, resulting in improved electrical
conductivity (10
2
-10
3
/square
13,32
) of these graphene
sheets, with structure more or less like those graphene
molecules.
26
And more importantly for its magnetic proper-
ties, vacancies and topological defects with unpaired elec-
trons would be introduced on the graphene sheets during this
process of annealing, due to the removal of the functional
groups, such as carbon dioxide, OH, etc.
18,32
The removal
of these groups under certain conditions would introduce
various edge and defect sites with spin units on sp
2
carbon
materials, and this has been studied thoroughly and elegantly
by Radovic.
18
These defects could include carbene, benzye,
18
phenylene units,
17
topological defects,
16,18,32
or a more
complicated zigzag edge.
18
It should be pointed out that both
theoretical
4,7,8,18
and experimental
4,7
works have proved that
most if not all the above unit structures with net spin are
stable within a large conjugation system such as in graphene
at room temperature. We want further to emphasize that
many stable molecules with structures the same or similar
as those proposed spin units
15-18,32
in graphene have been
isolated and characterized fully mainly by Haddons
group,
20-22
and the reason for their stability is due to the
huge -conjugation in these molecules. Then, if indeed the
long-range orderly magnetically coupling of these spins
exists, for example with the PKKY mechanism
15
via either
intramolecular interaction in individual graphene sheets or/
and intermolecular/lattice interaction between neighboring
graphene sheets,
9,15-19,21,33
the ferromagnetism could then
arise. While results are based on current theoretical work
and experimental results, we believe that the room-temper-
ature ferromagnetism is an intrinsic property of graphene-
based materials, and we are looking for direct and conclusive
evidence, such as some topographic/structural data. Further-
more, the conrmation and understanding for this pure
organic material room-temperature ferromagnetism, as one
of the most difcult topics in physics and material science,
certainly deserve much more through study, and the results
could be very rewarding with far and wide implications.
In summary, we have observed the room-temperature
ferromagnetism of bulk graphene materials prepared from
soluble functionalized graphene sheets. The possible origin
of the ferromagnetism may come from the long-range
coupling of spin units existing as defects in graphene sheets,
which are generated in the annealing process. But more direct
evidence is needed for a conclusion. The likely intrinsic
room-temperature ferromagnetism, combined with its semi-
Figure 2. An enlarged M vs H loop at T ) 300 K. (a) Sample Graphene-400, hysteresis was observed in the eld range -2500 Oe < H
< 2500 Oe, H
c
) 46 Oe. (b) Sample Graphene-600, hysteresis was observed in the eld range -1500 Oe < H < 1500 Oe, H
c
) 40 Oe.
Figure 3. Magnetization hysteresis loops at 2 K. (a) Sample
Graphene-400 (0), M
s
) 0.25 emu/g. (b) Sample Graphene-600
(2), M
s
) 0.90 emu/g.
Figure 4. An enlarged M vs H loop at T ) 2 K. (a) Sample Graphene-400, hysteresis observed in the eld range -2500 Oe < H < 2500
Oe, H
c
) 33 Oe. (b) Sample Graphene-600, hysteresis observed in the eld range -6000 Oe < H < 6000 Oe, H
c
) 596 Oe.
222 Nano Lett., Vol. 9, No. 1, 2009
conductivity and solution-processing capability, should have
wide-reaching implications in material science, and these
collective properties could make graphene-based materials
a competent choice for many important device applications,
including spintronic, magnetoresistance, magnetic memory
devices, and so on. Our results may also prompt more
thorough and carefully studies for the room-temperature
ferromagnetism of pure carbon-based materials, including
C
60
and carbon nanotubes.
Methods. Materials. Graphene oxide (GO) was synthe-
sized by the modication of Hummers method
13,25
and
described elsewhere.
13
GO was placed in a vacuum desic-
cator, where a piece of lter paper saturated with the
chemical reducing agent hydrazine (80% water solution) was
placed for the rst step of reduction. The reducing reaction
time was controlled for 24 h to generate the partially reduced
graphene oxide. The same partially reduced graphene oxide
sample was then separated into batches of 100 mg and then
annealed in argon atmosphere at 400, 600, and 800 C to
get the samples of Graphene-400, Graphene-600, and
Graphene-800, respectively. The heating rate from room
temperature to the desired temperature was 3 C/min, and
then the temperature was kept at the desired temperature for
3 h and nally cooled down to room temperature. The
annealing and cooling processes were strictly applied under
argon atmosphere.
Instruments. Magnetization measurements were performed
using a Quantum Design MPMS XL-7 superconducting
quantum interference device (SQUID) magnetometer. The
powder sample of graphene was packed in a bit of diamag-
netic plastic lm, and then the packed sample was put in a
diamagnetic plastic straw and impacted into a minimal
volume for magnetic measurements. Background magnetic
measurements were checked for the packing material. At 2
or 300 K, the sample was installed in the SQUID chamber
and the eld dependence of magnetization was measured.
The magnetic properties of the pristine partially reduced
graphene oxide sample before annealing were also measured
with the samples of Graphene-400, Graphene-600, and
Graphene-800 for comparison. Electric properties were
measured at 300 K using a QUANTUM DESIGN PPMS-9.
Aqueous solutions of graphene oxide (10 mg/mL) were spin-
coated onto a glass slide (500 revolutions/min, 12 s; 1000
revolutions/min, 30 s), and subsequently the lm was carried
out with the hydrazine reduction and annealing procedure
strictly according to the same procedures for the preparation
of Graphene-400/600/800 samples. Au electrodes were then
vacuum deposited on the lm, and then its conductivity was
measured using a standard four-probe method. Impurity
elements (Fe, Co, and Ni) of the pristine partially reduced
graphene oxide were determined by atomic absorption
spectrophotometer (Hitachi 180-80). The AAS analysis
samples were prepared by dissolving samples completely
using concentrated nitric acid at 100 C.
Acknowledgment. We gratefully acknowledge the -
nancial support from the NSFC (#20774047), MoST
(#2006CB932702) of China, and NSF of Tianjin City
(#07JCYBJC03000 and 08JCZDJC25300).
References
(1) Miller, J. S. Organometallic- and organic-based magnets: New
chemistry and new materials for the new millennium. Inorg. Chem.
2000, 39, 43924408.
(2) Tamura, M.; Nakazawa, Y.; Shiomi, D.; Nozawa, K.; Hosokoshi, Y.;
Ishikawa, M.; Takahashi, M.; Kinoshita, M. Bulk ferromagnetism in
the -phase crystal of the p-nitrophenyl nitronyl nitroxide radical.
Chem. Phys. Lett. 1991, 186, 401404.
(3) Allemand, P. M.; Khemani, K. C.; Koch, A.; Wudl, F.; Holczer, K.;
Donovan, S.; Gruner, G.; Thompson, J. D. Organic molecular soft
ferromagnetism in a fullerene C
60
. Science 1991, 253, 301303.
(4) Han, K. H.; Spemann, D.; Esquinazi, P.; Hohne, R.; Riede, V.; Butz,
T. Ferromagnetic spots in graphite produced by proton irradiation.
AdV. Mater. 2003, 15, 17191722.
(5) Makarova, T. Magnetism of carbon-based materials: Studies of high-
Tc superconductivity; NOVA Science Publishers, Inc.: New York,
2003; Vol. 44.
(6) Makarova, T. L.; Sundqvist, B.; Hohne, R.; Esquinazi, P.; Kopelevichk,
Y.; Scharff, P.; Davydov, V. A.; Kashevarova, L. S.; Rakhmanina,
A. V. Magnetic carbon. Nature 2001, 413, 716718.
(7) Esquinazi, P.; Spemann, D.; Hohne, R.; Setzer, A.; Han, K.-H.; Butz,
T. Induced magnetic ordering by proton irradiation in graphite. Phys.
ReV. Lett. 2003, 91, 227201.
(8) Service, R. F. American physical Society meeting: A positive spin on
semiconductors. Science 2004, 304, 4243.
(9) Harigaya, K. The mechanism of magnetism in stacked nanographite:
theoretical study. J. Phys.: Condens. Matter 2001, 13, 12951302.
(10) Lee, H.; Son, Y. W.; Park, N.; Han, S. W.; Yu, J. J. Magnetic ordering
at the edges of graphitic fragments: Magnetic tail interactions between
the edge-localized states. Phys. ReV. B 2005, 72, 174431.
(11) Novoselov, K. S.; Geim, A. K.; Morozov, S. V.; Jiang, D.; Zhang,
Y.; Dubonos, S. V.; Grigorieva, I. V.; Firsov, A. A. Electric eld effect
in atomically thin carbon lms. Science 2004, 306, 666669.
(12) Gomez-Navarro, C.; Burghard, M.; Kern, K. Elastic properties of
chemically derived single graphene sheets. Nano Lett. 2008, 8, 2045
2049.
(13) Becerril, H. A.; Mao, J.; Liu, Z. F.; Stoltenberg, R. M.; Bao, Z. N.;
Chen, Y. S. Evaluation of solution-processed reduced graphene oxide
lms as transparent conductors. Acs Nano 2008, 2, 463470.
(14) Novoselov, K. S.; Geim, A. K.; Morozov, S. V.; Jiang, D.; Katsnelson,
M. I.; Grigorieva, I. V.; Dubonos, S. V.; Firsov, A. A. Two-
dimensional gas of massless Dirac fermions in graphene. Nature 2005,
438, 197200.
(15) Vozmediano, M. A. H.; Lopez-Sancho, M. P.; Stauber, T.; Guinea,
F. Local defects and ferromagnetism in graphene layers. Phys. ReV.
B 2005, 72, 155121.
(16) Yazyev, O. V.; Helm, L. Defect-induced magnetism in graphene. Phys.
ReV. B 2007, 75, 125408.
(17) Wang, W. L.; Meng, S.; Kaxiras, E. Graphene nanoakes with large
spin. Nano Lett. 2008, 8, 241245.
(18) Radovic, L. R.; Bockrath, B. On the chemical nature of graphene edges:
Origin of stability and potential for magnetism in carbon materials.
J. Am. Chem. Soc. 2005, 127, 59175927.
(19) Fernandez-Rossier, J.; Palacios, J. J. Magnetism in graphene nanois-
lands. Phys. ReV. Lett. 2007, 99, 177204.
(20) Itkis, M. E.; Chi, X.; Cordes, A. W.; Haddon, R. C. Magneto-opto-
electronic bistability in a phenalenyl-based neutral radical. Science
2002, 296, 14431445.
(21) Haddon, R. Design of organic metals and superconductors. Nature
1975, 256, 394396.
(22) Beer, L.; Mandal, S. K.; Reed, R. W.; Oakley, R. T.; Tham, F. S.;
Donnadieu, B.; Haddon, R. C. The First Electronically stabilized
phenalenyl radical: effect of substituents on solution chemistry and
solid-state structure. Cryst. Growth Des. 2007, 7, 802809.
(23) Stankovich, S.; Dikin, D. A.; Piner, R. D.; Kohlhaas, K. A.;
Kleinhammes, A.; Jia, Y. Y.; Wu, Y.; Nguyen, S. T.; Ruoff, R. S.
Synthesis of graphene-based nanosheets via chemical reduction of
exfoliated graphite oxide. Carbon 2007, 45, 15581565.
(24) Si, Y.; Samulski, E. T. Synthesis of water soluble graphene. Nano
Lett. 2008, 8, 16791682.
(25) Niyogi, S.; Bekyarova, E.; Itkis, M. E.; McWilliams, J. L.; Hamon,
M. A.; Haddon, R. C. Solution properties of graphite and graphene.
J. Am. Chem. Soc. 2006, 128, 77207721.
(26) Wu, J.; Pisula, W.; Mullen, K. Graphenes as potential material for
electronics. Chem. ReV. 2007, 107, 718747.
Nano Lett., Vol. 9, No. 1, 2009 223
(27) Liu, Z.; Liu, Q.; Huang, Y.; Ma, Y. F.; Yin, S. G.; Zhang, X. Y.;
Sun, W.; Chen, Y. S. Organic Photovoltaic devices based on a novel
acceptor material: Graphene. AdV. Mater. 2008, 20, 3924-3930.
(28) Hummers, W.; Offeman, R. Preparation of graphite oxide. J. Am.
Chem. Soc. 1958, 80, 1339.
(29) Li, D.; Muller, M. B.; Gilje, S.; Kaner, R. B.; Wallace, G. G.
Processable aqueous dispersions of graphene nanosheets. Nat. Nano-
technol. 2008, 3, 101105.
(30) Murata, K.; Ueda, H.; Kawaguchi, K. Preparation of carbon powders
by pyrolysis of cyclododecane under vacuum and their magnetic
properties. Synth. Met. 1991, 44, 357362.
(31) Gomez-Navarro, C.; Weitz, R. T.; Bittner, A. M.; Scolari, M.; Mews,
A.; Burghard, M.; Kern, K. Electronic transport properties of individual
chemically reduced graphene oxide sheets. Nano Lett. 2007, 7, 3499
3503.
(32) Schniepp, H. C.; Li, J. L.; McAllister, M. J.; Sai, H.; Herrera-Alonso,
M.; Adamson, D. H.; Prudhomme, R. K.; Car, R.; Saville, D. A.;
Aksay, I. A. Functionalized single graphene sheets derived from
splitting graphite oxide. J. Phys. Chem. B 2006, 110, 85358539.
(33) Rajca, A.; Wongsriratanakul, J.; Rajca, S. Magnetic ordering in an
organic polymer. Science 2001, 294, 15031505.
NL802810G
224 Nano Lett., Vol. 9, No. 1, 2009