Observation of the Goos-Hänchen shift in

graphene via weak measurements

Cite as: Appl. Phys. Lett. 110, 031105 (2017); https://doi.org/10.1063/1.4974212
Submitted: 03 November 2016 • Accepted: 05 January 2017 • Published Online: 18 January 2017

Shizhen Chen, Chengquan Mi, Liang Cai, et al.

ARTICLES YOU MAY BE INTERESTED IN

Goos–Hänchen effect enabled optical differential operation and image edge detection
Applied Physics Letters 116, 211103 (2020); https://doi.org/10.1063/5.0006483

Ultrasensitive and real-time detection of chemical reaction rate based on the photonic spin
Hall effect
APL Photonics 5, 016105 (2020); https://doi.org/10.1063/1.5131183

Wavelength-independent optical fully differential operation based on the spin–orbit

interaction of light
APL Photonics 5, 036105 (2020); https://doi.org/10.1063/1.5144953

Appl. Phys. Lett. 110, 031105 (2017); https://doi.org/10.1063/1.4974212 110, 031105

© 2017 Author(s).
 APPLIED PHYSICS LETTERS 110, 031105 (2017)

€nchen shift in graphene via weak measurements

Observation of the Goos-Ha
Shizhen Chen, Chengquan Mi, Liang Cai, Mengxia Liu, Hailu Luo,a) and Shuangchun Wen
Laboratory for Spin Photonics, School of Physics and Electronics, Hunan University, Changsha 410082,
China
(Received 3 November 2016; accepted 5 January 2017; published online 18 January 2017)
We report the observation of the Goos-H€anchen effect in graphene via a weak value amplifica-
tion scheme. We demonstrate that the amplified Goos-H€anchen shift in weak measurements is
sensitive to the variation of graphene layers. Combining the Goos-H€anchen effect with weak
measurements may provide important applications in characterizing the parameters of graphene.
Published by AIP Publishing. [http://dx.doi.org/10.1063/1.4974212]

The behavior of plane wave in reflection can be simply requirement of a very large weak value, the final output
predicted by geometrical optics. However, for the bounded intensity is still strong and measured data are stable. Our
beam of light, it may undergo extra shifts due to the occur- result suggests that this technique may become an alternative
rence of diffractive corrections. Such shifts are known as way to effectively and conveniently identify layers of few-
the Goos-H€anchen (GH)1 and the Imbert-Fedorov2,3 shifts layer graphene.
according to the directions parallel and perpendicular to the Consider a reflection system shown in Fig. 1. An incom-
plane of incidence, respectively. In recent years, the research ing beam is at the 45 linearly polarization state jAi. Under a
of the GH shift is still active although it was discovered more TIR condition, the reflected light consists of two beams, one
than 60 years ago. The corresponding studies reach not only is the horizontal (H) component jHi displaced by XH and the
to the investigation of the inherent physics behind this phe- other is the vertical (V) component jVi displaced by XV. We
nomenon4–8 but also to the behavior of the shift at various only concern about the difference between XV and XH as it is
reflecting surfaces.9–15 For example, in photonic crystals,16 the measured variable in our weak measurement scheme. To
the GH shift for self-collimated beams were experimental calculate the GH shift (XV – XH) for different graphene
observed.17 In the system with graphene, it is very interesting layers, the obtainment of the reflection coefficient in the
for the observation of beam shifts due to flexible optical prop- graphene-dielectric interface is important.
erties. The GH shift in reflection from a graphene-dielectric There are many models of reflection coefficients in gra-
interface is therefore extensively studied. The Fresnel reflec- phene. For example, in previous works the slab-based model
tion coefficients with the existence of graphene become dif- is used to calculate the beam shifts in multi-layer gra-
ferent,18,19 and the behavior of the GH shift is changed or phene.23,39 This model is deduced by taking the refractive
tunable.20,21 In particular, the GH shift on a substrate coated indexes of graphene into account in each independent
with graphene can be quantized in an external magnetic component. For monolayer graphene, based on the boundary
field;22 though in a common environment, a giant GH shift in conditions with related parameters, a nice model called
graphene was observed.23 For electron beams, there also exits Fresnel-based model is introduced by Merano to study the
the GH effect in reflection. The shift becomes giant in gra-
phene double-barrier structures,24 and an electron spin beam
splitter was realized by using this effect in graphene.25
In this letter, we investigate the GH shift in graphene
under a total internal reflection (TIR) condition. The GH
shift is still very small under this condition and difficult to be
directly observed. However, using an amplify technique
called weak measurements, such a difficulty can be over-
come. Weak measurement is an important and convenient
approach that has reached fruitful achievements for detecting
light beam shifts.26–32 This conception was first proposed in
the context of quantum mechanics and then has been exten-
sively studied.33–36 With the help of weak measurements, the
GH shift in TIR37 or in partial reflection38 was observed
recently. In our work, we find that the shift in graphene
amplified by a weak value scheme is sensitive to the gra-
phene layers. Because the GH effect occurs in the regime of
TIR (the reflected intensity is equal to incident intensity) and
FIG. 1. Schematic illustrating the GH shift in graphene under TIR. A 45
the success rate of the postselection is still large due to no linearly polarized beam labeled by jAi hits a glass-graphene-air interface. Its
components jHi and jVi experience different lateral small GH shifts. The
a) green arrows indicate the polarization direction.
Electronic mail: hailuluo@hnu.edu.cn

0003-6951/2017/110(3)/031105/5/$30.00 110, 031105-1 Published by AIP Publishing.

031105-2 Chen et al. Appl. Phys. Lett. 110, 031105 (2017)

optical behavior in graphene.40,41 Our work uses the slab

model since we now deal with the problem in multi-layer
graphene. The reflection coefficient is given by

RA þ R0A expð2idkgz Þ
rA ¼ : (1)
1 þ RA R0A expð2idkgz Þ

Here, RA and R0A are the Fresnel reflection coefficients in the

glass-graphene and graphene-air interfaces, respectively, A 2
{H, V}. kgz is the component of the wave vector k0 in gra-
phene along the z direction. d ¼ mDd is the thickness of the
graphene film, with m and Dd ¼ 0.34 nm representing the
layer numbers and the thickness of a single layer graphene,
respectively. In TIR, the reflection coefficient rA is complex
and can be written as rA ¼ jrA j expðiuA Þ. We calculate the
spatial and angular GH shifts from the reflection coefficients.
The expression of GH shifts for H and V components can be
written as4,15,18,19,42

1 @uA z 1 @jrA j
XA ¼  ; (2)
nk0 @hi zr nk0 jrA j @hi

where n ¼ 1.515 is the refractive index of prism, zr is the

Rayleigh range, z is the propagation, and hi is an incident
angle. We chose the refractive indexes of graphene as FIG. 2. (a) (XV-XH) as a function of incident angle for different graphene
layers. The red dashed line indicates the incident angle (42 ) in our experi-
(3 þ 1.149i).39,43 The first term and the second term in Eq. ment, and the angle of total reflection is 41.3 . (b) Representation on the
(2) are, respectively, related to the spatial and angular shifts. Poincare sphere of the states jii; jcf i, and jf i. jii is the incident state and jf i
From Eq. (2), we plot the curves of the GH shift in Fig. 2(a) is the postselected state. jcf i is the preselected state after reflecting at the
graphene-dielectric interface. The state jcf i is off to be antiparallel to jii in
as a function of incident angle for different graphene layers.
two angular directions depending on the graphene layer numbers. The red
The shift decreases with increasing numbers of layer, but the dashed circle represents the trajectory of jf i when we rotate the half-wave
decrement is small. For a specific case, we, respectively, cal- plate (HWP). (c) Experimental setup: a Gaussian beam generated by a
culate the angular and spatial shifts for monolayer graphene He-Ne laser (632.8 nm). GLP1 and GLP2, Glan laser polarizers; QWP,
quarter-wave plate; L1 and L2 lenses with focal length 125 mm and
at 42 , and they are about 130.8 nm and 880 nm. In particu- 250 mm, respectively. The beam waist w after L1 is 71.25 lm. The data are
lar, using the Fresnel-based model, the results are almost the detected by a CCD. Insets show the rotations of GLP1, QWP, and GLP2.
same (142.4 nm and 900.4 nm).  
To amplify these small shifts, the weak measurements ^ ¼ DH
GH
0
; (4)
are employed. The experimental setup is plotted in Fig. 2(c), 0 DV
which is similar to that for observing the GH effect without where
graphene.37 In a schema of weak measurements, the prese-
lected state, postselected state, and weak coupling between 1 @uA 1 @jrA j
the system and pointer are three key elements. The corre- DA ¼ i : (5)
nk0 @hi nk0 jrA j @hi
sponding elements in our scheme will be clear in the follow-
ing paragraph. We next analyze the postselection, which can be realized by
We first discuss the preselected state which is delicate the combination of quarter-wave plate (QWP), half-wave
due to the effect of reflection coefficients. In the experiment, plate (HWP), and GLP2. In the experiment, the optical axis

we set the optical axis of GLP1 to ffi to
pffiffi45 pffiffiproject
ffi the incident of QWP is fixed to 45 from the x axis, and we assume that
polarization state on jii ¼ ð1= 2; 1= 2Þ. And this state the rotation angle of HWP is a. These settings described by
changes to jcm i ¼ Fjii when the light is reflected at the inter- the Jones matrices are
face. F is the reflection matrix,44 " #
  1 1 i
rH 0 QWP ¼ pffiffiffi
F¼ ; (3) 2 i 1
0 rV " #
cosð2aÞ sinð2aÞ
of which action is also a part of the preselection process. The HWP ¼ : (6)
sinð2aÞ cosð2aÞ
preselected state for our weak measurement scheme is jcm i.
jcm i is an elliptical polarization state and different for one, The optical axis of GLP2 is (45 6 D) to project on the state
two, and three graphene layers. We next consider the weak
 
coupling. The tiny GH effect is regarded as a weak measur- hGLP2j ¼ cosð45 6DÞ sinð45 6DÞ : (7)
ing process here, as labeled by the dashed circle in Fig. 2(c).
In the language of quantum mechanics, this effect described Putting Eqs. (6) and (7) together, we obtain the postselected
by operator is6,38,42 state as
031105-3 Chen et al. Appl. Phys. Lett. 110, 031105 (2017)

  
hf j ¼ eið2a7DÞ eið2a7DÞ : (8) 1 z
Re½Aw ðDH  DV Þ þ Im½Aw ðDH  DV Þ : (11)
2 zr
For convenience, we represent the above states jii, jcm i, and
jf i on the Poincare sphere, as shown in Fig. 2(b). One can We have used a quantum mechanical description to ana-
see that the postselected state jf i is limited on the red dashed lyze the amplified GH shift in order to provide a good physi-
circle by adjusting HWP [from Eq. (8)]. For the preselected cal insight and simplify the analysis. However, the above
state jcm i, it exhibits a deviation from the red dashed circle theory will be invalid when the preselected and postselected
on the Poincare sphere in Fig. 2(b) [see Fig. 3(a) for the states are nearly orthogonal. Therefore, in order to obtain a
cases of different layers]. The reason for this deviation is more precise analysis for the weak measurement of GH
that jrH j 6¼ jrV j, and the difference between jrH j and jrV j effect, we next describe it by using standard wave optics.37
increases with increasing layers of graphene. We begin with the wave function of the beam after passing
With the preselected and postselected states discussed GLP1
above, the weak value in our weak measurements can be " #
2 2
obtained as x þ y
winc / exp iz  k0  ðx^ þ y^Þ; (12)
2ðzr þ izÞ
^ mi 1
hf jGHjc Aw
¼ ðD H þ D V Þ þ ðDH  DV Þ; (9)
hf jcm i 2 2 with the optical axis of GLP1 at 45 . When the beam reflects
from the glass-graphene-air interface, the function becomes4
where Aw ¼ hf jr^3 jcm i=hf jcm i, and r^3 is the Pauli matrix.
" #
The weak value from Eq. (9) is related to m, i.e., the layer of
ref x2 þ y2
graphene. Here, Aw in all cases is a complex number except w / exp iz  k0
2ðzr þ izÞ
the one m ¼ 0, in which it is pure imaginary. Note that the  

GH shift in TIR with graphene contains both spatial and the k0 x @ ln rH
 x^ rH 1  i
angular components, and the imaginary part of Aw can natu- zr þ iz @hi
rally convert the relevant spatial shift into an angular one. 

k0 x @ ln rV
Therefore, in order to obtain the centroid position of the þ y^ rV 1  i ; (13)
zr þ iz @hi
beam, the propagation effect in all cases should be consid-
ered.4,26 Containing Eq. (9), we obtain the amplified GH and we have ignored the terms responsible for the transverse
shift hxqua i as shift. Then QWP and HWP are introduced, and their actions
! ! are represented by Eq. (6). To obtain a state orthogonal to
^ mi
hf jGHjc z ^ mi
hf jGHjc the one in Eq. (7) as close as possible, HWP should be set as
Re þ Im : (10)
hf jcm i zr hf jcm i a ¼ ð3p þ uÞ=4, with u ¼ uH  uV . This is the key step to
obtain the minimum output intensity, which will be intro-
In fact, the shift we measure in the experiment is a relative duced later. After considering the operation of GLP2, the
position of the beams on CCD for the postselected states final wave function follows that
with 6D. Thus, after the irrelative terms in Eq. (10) are " #
neglected, the amplified GH shift hxqua i is given by x 2
þ y 2
wfin / exp iz  k0
2ðzr þ izÞ



p p
 cos 6D ð# þ ieiu Þ þ sin 6D ði# þ eiu Þ ;
4 4
(14)

 ;
FIG. 3. (a) Representation on the Poincare sphere of the state jcm i and jf i
when the tunable state jf i is closest orthogonal to jcm i (reading intensity on
CCD becomes minimum). The yellow area indicates the deviation to the cir-
cle of jf i. (b) The theoretical minimum intensity by adjusting HWP. (c) The
corresponding minimum intensity we read out from CCD. Note that the first,
second, third, and fourth columns correspond to the cases of no, one-layer,
two-layer, and three-layer graphene, respectively.
k0 x @ ln rV
0 x @ ln rH
where # ¼ rH 1  i zrkþiz @hi and  ¼ r V 1  i zr þiz @hi .
Now the amplified shift (the beam centroid) can be straight-
ÐÐ ÐÐ
forwardly calculated by hxcla i¼ xjwfin j2 dxdy= jwfin j2 dxdy
as
h i
2 jrH j2 ðzq  zr vÞ þ jrV j2 ðzr  zr sÞ þ njrH jjrV j
2k0 zr jrH j2 þ jrV j2 þ jrH j2 ðv2 þ q2 Þ þ jrV j2 ðr2 þ s2 Þ  1
(15)
where n ¼ cosð62DÞ½zr ðv þ sÞ  zðq þ rÞ  sinð62DÞ½zr
ðr  qÞ þ zðs  vÞ; 1 ¼ 2jrH jjrV j½cosð62DÞð2k0 zr þ qr þ vsÞ
þ sinð62DÞðqs  vrÞ; q ¼ Reð@ ln rH =@hi Þ; v ¼ Imð@ ln rH =
@hi Þ; r ¼ Reð@ ln rV =@hi Þ, and s ¼ Imð@ ln rV =@hi Þ. Equation
(15) indicates a large beam shift with a small angle D.
In the procedure to experimentally observe the amplified
GH shift, we first set GLP2 to 45 , i.e., D ¼ 0. Then we rotate
031105-4 Chen et al.
the HWP and the state jf i will project on the red dashed cir-
cle in Fig. 2(b). We adjust the HWP until the output intensity
on CCD becomes minimum. This indicates that the postse-
lected state jf i is most closely orthogonal to the preselected
state jcm i. We simulate the minimum intensity for different
graphene layers by I ¼ jwfin ðx; y; z; D ¼ 0Þj2 from Eq. (14).
In Fig. 3(b), we see that only in the case of no graphene the
minimum intensity exhibits a double-peak profile, which is a
usual distribution of the minimum output intensity in weak
measurements.34 This is because in that case the state jf i is
possible to be orthogonal to jc0 i. For other cases with the
existence of graphene, the tunable postselected state can not
be orthogonal to the preselected state, and the nonorthogonal
degree becomes larger when the layers of graphene increase,
leading to a smaller weak value. As a result, the minimum
intensity tends to a Gaussian form.45 The minimum intensity
in the experimental observation is shown in Fig. 3(c).
After minimizing the output intensity by adjusting
HWP, we rotate the GLP2 first to (45 þ D) and then to
(45 – D) to measure the final amplified shift. We measure
the shifts in no, one-layer, two-layer, and three-layer gra-
phene. The theoretical amplified shift is precisely obtained
from Eq. (15), and one can calculate the corresponding
amplified factor by hxcla i=ðXH  XV Þ. We plot the related
curves in Figs. 4(a) and 4(b) based on the slab-based model
of reflection coefficients. We see that the amplified factors
are not very large. But for a fixed angle D, the amplified shift
obviously decreases with increasing layers.
As we have introduced above, there are another model
of reflection coefficients (the Fresnel-based model) that can
be used to analyze the GH effect. In Fig. 4(c), we simulate
the theoretical curves according to this model, together with
the experimental data. For one-layer graphene at wavelength
633 nm, we choose the surface conductivity and the surface
susceptibility values as 6.08  105 X and 1.0  109 m,
respectively.40 Assuming that an individual graphene sheet
Appl. Phys. Lett. 110, 031105 (2017)
is a non-interacting monolayer, for few-layer graphene, the
surface conductivity and the surface susceptibility values
increase linearly with layer number m (m 6 5).46 So we
obtain the parameters for multi-layer graphene as m  6.08
 105 X and m  1.0  109 m. From Figs. 4(a) and 4(c),
we see that the predictions from the Fresnel-based model
and the slab-based model both consist of the experimental
results. For a better comparison between the Fresnel-based
model and the slab-based model, the discussion with various
experimental data is required.40,47,48
In our experiment, each sample we fabricate is uniform
layer and the size of sample is 1 cm  1 cm. In practice, we
repeat the experiment of weak measurements several times
in different measuring places for each sample and the data
are nearly the same. To confirm the corresponding layers of
graphene film, we provide their Raman spectra in Fig. 4(d).
The layers of each sample deduced from our data coincide
well with the results from Raman spectra.
The measurability of very small displacements is ulti-
mately limited by the quantum noise of the light, because
enough photons need to be collected to resolve the position
of the field.26 Note that another interesting beam shift
induced by the photonic spin Hall effect can also be used to
identify graphene layers.39 In that case, the experimental
measurement was performed near Brewster angle. Therefore,
the experimental data are a little unstable due to a low reflec-
tion intensity near Brewster angle. In the present case, how-
ever, enough photons can be captured by the detector due to
the TIR. From the error bars in Fig. 4, we see that the data
read from CCD are very stable. We recently noted that a sim-
ilar experimental setup can also be used to observe the angu-
lar and lateral GH shifts near the critical angle for TIR.49
In conclusion, we have experimentally observed the GH
shift in graphene via weak measurements. Theoretically, the
variations of the initial shifts for no, one-layer, two-layer,
and three-layer graphene are tiny. However, employing a
FIG. 4. (a) The amplified shifts calcu-
lated by the slab-based model as the
function of angle D for different gra-
phene layers from Eq. (15). (b) The
corresponding amplified factors. (c)
The amplified shift based on the
Fresnel-based model of reflection coef-
ficients for different graphene layers.
Experimental data are shown as open
dots with error bars. (d) Raman spec-
trum of our samples for one-layer,
two-layer, and three-layer graphene.
031105-5 Chen et al. Appl. Phys. Lett. 110, 031105 (2017)

24
weak value amplification scheme, the amplified GH shift Y. Song, H.-C. Wu, and Y. Guo, Appl. Phys. Lett. 100, 253116 (2012).
25
decreases with layer number of graphene. This technique Q. Zhang and K. S. Chan, Appl. Phys. Lett. 105, 212408 (2014).
26
O. Hosten and P. Kwiat, Science 319, 787 (2008).
may be utilized to identify layers of few-layer graphene with 27
P. B. Dixon, D. J. Starling, A. N. Jordan, and J. C. Howell, Phys. Rev.
the stable data of this system. Our research is important for Lett. 102, 173601 (2009).
28
the research of graphene in future and may offer the opportu- Y. Qin, Y. Li, H. He, and Q. Gong, Opt. Lett. 34, 2551 (2009).
29
Y. Gorodetski, K. Y. Bliokh, B. Stein, C. Genet, N. Shitrit, V.
nity to characterize the parameters of graphene with the help
Kleiner, E. Hasman, and T. W. Ebbesen, Phys. Rev. Lett. 109,
of weak measurements. 013901 (2012).
30
L.-J. Kong, X.-L. Wang, S.-M. Li, Y. Li, J. Chen, B. Gu, and H.-T. Wang,
This research was supported by the National Natural Appl. Phys. Lett. 100, 071109 (2012).
31
J. Ren, Y. Li, Y. Lin, Y. Qin, R. Wu, J. Yang, Y.-F. Xiao, H. Yang, and Q.
Science Foundation of China (Grant Nos. 11274106 and Gong, Appl. Phys. Lett. 101, 171103 (2012).
11474089). 32
X. Zhou, Z. Xiao, H. Luo, and S. Wen, Phys. Rev. A 85, 043809
(2012).
1 33
F. Goss and H. H€anchen, Ann. Phys. 436, 333 (1947). Y. Aharonov, D. Z. Albert, and L. Vaidman, Phys. Rev. Lett. 60, 1351
2
C. Imbert, Phys. Rev. D 5, 787 (1972). (1988).
3 34
F. I. Fedorov, J. Opt. 15, 014002 (2013) [Dokl. Akad. Nauk SSSR 105, I. M. Duck, P. M. Stevenson, and E. C. G. Sudarshan, Phys. Rev. D 40,
465 (1955) (in Russian)]. 2112 (1989).
4 35
A. Aiello and J. P. Woerdman, Opt. Lett. 33, 1437 (2008). A. G. Kofman, S. Ashhab, and F. Nori, Phys. Rep. 520, 43 (2012).
5 36
M. R. Dennis and J. B. G€ otte, New J. Phys. 14, 073013 (2012). J. Dressel, M. Malik, F. M. Miatto, A. N. Jordan, and R. W. Boyd, Rev.
6
F. T€oppel, M. Ornigotti, and A. Aiello, New J. Phys. 15, 113059 (2013). Mod. Phys. 86, 307 (2014).
7 37
K. Y. Bliokh and A. Aiello, J. Opt. 15, 014001 (2013). G. Jayaswal, G. Mistura, and M. Merano, Opt. Lett. 38, 1232 (2013).
8 38
M. Ornigotti, A. Aiello, and C. Conti, Opt. Lett. 40, 558 (2015). S. Goswami, S. Dhara, M. Pal, A. Nandi, P. K. Panigrahi, and N. Ghosh,
9
H. M. Lai and S. W. Chen, Opt. Lett. 27, 680 (2002). Opt. Express 24, 6041 (2016).
10 39
H. Gilles, S. Girard, and J. Hamel, Opt. Lett. 27, 1421 (2002). X. Zhou, X. Ling, H. Luo, and S. Wen, Appl. Phys. Lett. 101, 251602
11
D.-K. Qing and G. Chen, Opt. Lett. 29, 872 (2004). (2012).
12 40
X. Yin, L. Hesselink, Z. Liu, N. Fang, and X. Zhang, Appl. Phys. Lett. 85, M. Merano, Phys. Rev. A 93, 013832 (2016).
41
372 (2004). M. Merano, Opt. Lett. 41, 5780 (2016).
13 42
X. Yin and L. Hesselink, Appl. Phys. Lett. 89, 261108 (2006). G. Jayaswal, G. Mistura, and M. Merano, Opt. Lett. 39, 6257 (2014).
14 43
J. He, J. Yi, and S. He, Opt. Express 14, 3024 (2006). M. Bruna and S. Borini, Appl. Phys. Lett. 94, 031901 (2009).
15 44
M. Merano, A. Aiello, G. W.’t Hooft, M. P. van Exter, E. R. Eliel, and J. B. G€otte, W. L€offler, and M. R. Dennis, Phys. Rev. Lett. 112, 233901
J. P. Woerdman, Opt. Express 15, 15928 (2007). (2014).
16 45
X. Yu and S. Fan, Appl. Phys. Lett. 83, 3251 (2003). S. Chen, X. Zhou, C. Mi, H. Luo, and S. Wen, Phys. Rev. A 91, 062105
17
A. Matthewsa and Y. Kivshar, Appl. Phys. Lett. 93, 131901 (2008). (2015).
18 46
S. Grosche, M. Ornigotti, and A. Szameit, Opt. Express 23, 30195 (2015). H. Yan, X. Li, B. Chandra, G. Tulevski, Y. Wu, M. Freitag, W. Zhu, P.
19
N. Hermosa, J. Opt. 18, 025612 (2016). Avouris, and F. Xia, Nat. Nanotechnol. 7, 330 (2012).
20 47
J. C. Martinez and M. B. A. Jalil, Europhys. Lett. 96, 27008 (2011). P. Blake, E. W. Hill, A. H. Castro Neto, K. S. Novoselov, D. Jiang, R.
21
M. Cheng, P. Fu, X. Chen, X. Zeng, S. Feng, and R. Chen, J. Opt. Soc. Yang, T. J. Booth, and A. K. Geim, Appl. Phys. Lett. 91, 063124
Am. B 31, 2325 (2014). (2007).
22 48
W. J. M. Kort-Kamp, N. A. Sinitsyn, and D. A. R. Dalvit, Phys. Rev. B 93, V. G. Kravets, A. N. Grigorenko, R. R. Nair, P. Blake, S. Anissimova,
081410(R) (2016). K. S. Novoselov, and A. K. Geim, Phys. Rev. B 81, 155413 (2010).
23 49
X. Li, P. Wang, F. Xing, X.-D. Chen, Z.-B. Liu, and J.-G. Tian, Opt. Lett. O. J. S. Santana, S. A. Carvalho, S. De Leo, and L. E. E. De Araujo, Opt.
39, 5574 (2014). Lett. 41, 3884 (2016).