Detection of graphene chirality using achiral liquid crystalline platforms

Rajratan Basu,a) Daniel Kinnamon, and Alfred GarveySoft Matter and Nanomaterials Laboratory, Department of Physics, The United States Naval Academy,Annapolis, Maryland 21402, USA

(Received 21 July 2015; accepted 5 September 2015; published online 16 September 2015)

Monolayer graphene flakes were dispersed at low concentrations into two achiral liquid crystals

(LCs) alkoxyphenylbenzoate (9OO4) and 4-cyano-40-pentylbiphenyl (5CB), separately. The

presence of graphene resulted in two types of chiral signatures in the LCs: an electroclinic effect

(a polar tilt of the LC director perpendicular to, and linear in, an applied electric field) in the

smectic-A phase of 9OO4, and a macroscopic helical twist of the LC director in the nematic phase

of 5CB. Graphene flakes generally possess strain chirality and edge chirality. The non-covalent

interactions between the LC molecules and chiral graphene flakes induce molecular conformational

deracemization in the LC, exhibiting a bulk electroclinic effect and a macroscopic helical twist.

[http://dx.doi.org/10.1063/1.4931147]

I. INTRODUCTION

Chirality is one of the most important and interesting

topics in science. In nanoscience, carbon nanotubes are a

prime example of chiral macromolecule. Apart from carbon

nanotubes, it has been recently demonstrated that graphene

also possesses chirality.1–6 Graphene is a crystalline allo-

trope of carbon with 2-dimensional properties. The carbon

atoms in a graphene sheet are densely packed in a regular

sp2-bonded atomic-scale hexagonal pattern. An infinitely

large graphene sheet is not expected to show any chirality

due to the symmetry of the hexagonal pattern. However,

when graphene is present in form of nano flakes, the symme-

try breaking of the graphene lattice due to a soft shear mode

in the graphene sheet shows circular dichroism,1,2 indicating

the presence of surface chirality. Apart from this strain chir-ality, graphene also exhibits edge chirality. Due to the pres-

ence of different chiralities, i.e., armchair or zigzag at the

edge,3–6 the flakes show different optical,7 superconductive,8

and magnetic properties.9 This edge chirality of graphene

can be detected from the intensity of the D band using

Raman spectroscopy.10,11 In soft matter, chirality is an

important aspect of the phases, as the component molecules

often can exist both in chiral and achiral form. In rod-like

liquid crystals (LC), the introduction of chirality in the mole-

cules breaks the mirror symmetry and shows unique features

in chiral LC phases, such as macroscopic director twist in

the chiral nematic, electroclinic effect in the chiral smectic-

A, and ferroelectricity in the chiral smectic-C. Thus, the pres-

ence of chirality in soft matter results in a wide range of

physical properties, from electro-optical to hierarchical and

ultimately macroscopic helical twist.

In a different direction, two-dimensional honeycomb

structure of graphene shows several interesting interactions

with LCs. The graphene nanostructure can be used to enhance

the tilted smectic-C order.12 Multilayer graphene flakes can

improve the electro-optic response in a nematic phase.13 The

vertical alignment of LC molecules can be achieved by

graphene-oxide without any surface treatment of the sub-

strate.14 Graphene with benzene can enhance the orientational

order in a nematic phase.15 Functionalized graphene can

mutually self-assemble discotic liquid crystals.16 Transparent

graphene-conducting-layers, instead of indium-tin-oxide, can

be used as electrodes to produce high-transmittance liquid

crystal displays.17,18 In an LC – nanocolloidal dispersion, the

effect of nonmesogenic guest-nanoparticles on the LC’s bulk

properties often rests on the molecular identification at the

nanoscale in order to share and disseminate the “information”

coded into the nanostructure of the nanoparticles.19–22

Because most of the graphene flakes can have strain chirality,1

we ask the question: Can this surface chirality be transmitted

into the LC? If so, the resulting LC-graphene mixture could

exhibit a spatially averaged bulk chirality. Here, we report ex-

perimental observations of chirality transfer from graphene

surface into two different achiral LCs, alkoxyphenylbenzoate

(9OO4) and 4-cyano-40-pentylbiphenyl (5CB). An electro-

clinic effect23 (a polar tilt of the LC director perpendicular to,

and linear in, an applied electric field) is observed in the

smectic-A phase of 9OO4 when the LC is doped with mono-

layer graphene flakes. On the other hand, a macroscopic

helical twist24 of the LC director is observed in the nematic

phase of 5CB in the presence of graphene flakes.

II. EXPERIMENTS, RESULTS, AND DISCUSSION

The LC 9OO4 (obtained from LC Vision) has the cooling

phase sequence as: Isotropic – 87� – nematic – 73.5� – smectic-

A – 62.5� – smectic-C – 50.2� – smectic-B – 35� – Crystal. The

LC 5CB has the cooling phase sequence as: Isotropic – 35� –

nematic – 22� – Crystal. The chemical structures of the two

achiral liquid crystals used in this experiment are shown in

Figure 1. The pristine graphene (GP) sample, obtained from

Graphene Supermarket, Inc., contained more than 97% of

monolayer flakes (and 3% of multilayer flakes) of an average

thickness of 0.35 nm and an average lateral size of 550 nm. The

GP sample was ultrapure with no oxidation, no surfactants, and

99.99% carbon content. The GP sample was available in etha-

nol solvent in 1 mg/l. The ethanol þ GP solution was first

remixed by a micro-homogenizer tip of 5 mm diameter at

a)Author to whom correspondence should be addressed. Electronic mail:

basu@usna.edu.

0021-8979/2015/118(11)/114302/5/$30.00 118, 114302-1

JOURNAL OF APPLIED PHYSICS 118, 114302 (2015)

[This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to ] IP:

73.163.157.169 On: Wed, 16 Sep 2015 14:35:50

35 000 rpm for 3 h, followed by sonication for 10 h. The LC

then was added to the ethanol þ GP mixture and sonicated for

5 h, allowing the LC to dissolve completely into the solution.

The ethanol was evaporated slowly at an elevated temperature,

leaving a pure LC þ GP mixture. Finally, the LC þ GP mix-

ture was degassed under a vacuum for 1 h. The final concentra-

tion of 9OO4 þ GP mixture was 0.0035 wt. % and the same

for the 5CB þ GP mixture was 0.0025 wt. %. For consistency,

the pure LCs were also treated the same way, such as dissolv-

ing in ethanol followed by a slow evaporation and degassing.

An electroclinic effect is observed in a bulk chiral smec-

tic-A LC.23 In this effect, an electric field ~E applied parallel to

the smectic layer induces a polar tilt h [/ E] of the director in

a plane perpendicular to ~E. The tilt susceptibility, also known

as the electroclinic coefficient ec¼ dhdE, diverges on cooling to-

ward the chiral smectic-C* phase (the asterisk denotes the chi-

ral phase). The electroclinic effect involves the reduced C2

symmetry and is absent when the system is achiral or racemic.

Chiral-based effects at LC/substrate interfaces also have been

known for some time.25–28 In these systems, either the LC or

the interaction surface layer is chiral, and the interface’s

reduced symmetry gives rise to surface electroclinic behavior.

Commercially manufactured LC cells (LC2-5.0, planar

antiparallel rubbed, from Instec, Inc.) with a 1.5� pre-tilt

angle, 0.5� 0.5 cm2 semitransparent indium tin oxide (ITO)

coated area, and a d¼ 5 lm spacing were used for our elec-

troclinic experiment. Two cells were filled at temperature in

the isotropic phase by capillary action, one with 9OO4 only

and the other with the 9OO4þGP mixture; from then on

both cells always remained in the smectic-C phase (or higher

temperature). Before performing any electroclinic measure-

ments, the 9OO4þGP cell was examined using a polarizing

optical microscope, which revealed uniform nematic,

smectic-A, and smectic-C textures like that of the pure LC

cell, indicating a uniform director field.

The optical setup consisted of a beam from a 5-mW

He-Ne laser at wavelength 633 nm that passed through the

polarizer, the cell, a crossed analyzer, and into a photodetec-

tor (PDA100A, Thor Labs). The beam was polarized at an

angle of p/8 with respect to the cells’ rubbing direction; this

is the classical “electroclinic geometry.”29 The output of the

detector was fed into both a lock-in amplifier (Stanford

Research Systems SR830 DSP) that was referenced to the

driving frequency f of the applied electric field (Stanford

Research Systems SR345) and to a dc voltmeter (Keithley

2002), allowing us to measure the ac intensity Iac at fre-

quency f and the dc intensity Idc, respectively. The setup and

data acquisition were computer controlled using LabVIEWVR

software. From the measured intensities, the field-induced

spatially average tilt angle h was obtained using the formula

h ¼ Iac

4 Idc.29 This is a common technique to measure h vs. E (f)

for the electroclinic effect in chiral LCs.28,30–33 In mean field

theory, the field-induced tilt angle h ¼ kEðT�TAC� Þ,

23 where TAC*

is the smectic-A to smectic-C* transition temperature and kis a proportionality constant, which reflects the chiral cou-

pling to the electric field, and vanishes in the absence of

chirality.

Figure 2 shows the tilt angle h as a function of the rms

applied field E at f¼ 1 kHz for (a) pure 9OO4 and for (b) the

9OO4þGP mixture. As expected, 9OO4 in the absence of GP

does not exhibit any electroclinic effect in the smectic-A phase,

confirming the absence of molecular chirality associated with

FIG. 1. Chemical structures of 5CB and 9OO4. The schematic diagram

represents that two parallel graphene flakes in 5CB LC induce a helical twist

of the LC director due to p–p stacking. Note that the flakes show different

chiralities along the edges. The p–p electron stacking is illustrated by match-

ing 5CB LC’s benzene rings (red) on the graphene-honeycomb structure.

FIG. 2. Electroclinic effect in the smectic-A phase: (a) tilt angle h vs. E(f¼ 1 kHz) for bulk 9OO4 at two different values of T, listed in the legend;

(b) tilt angle h vs. E (f¼ 1 kHz) for 9OO4þGP at eight different values of Tlisted in the legend. Lines represent linear fits.

114302-2 Basu, Kinnamon, and Garvey J. Appl. Phys. 118, 114302 (2015)

[This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to ] IP:

73.163.157.169 On: Wed, 16 Sep 2015 14:35:50

pure 9OO4. On the other hand, the GP-doped 9OO4 clearly

shows a bulk electroclinic effect that grows on cooling toward

TAC*, demonstrating a net chirality in the smectic-A phase. We

also examined the temperature behavior of the electroclinic

effect of the mixture above TAC* (¼ 62 �C), where ec ¼ dhdE

was found to increase rapidly on cooling, with an apparent

divergence at TAC*. In Figure 3, we plot e�1c vs. T, observing

an approximately linear variation with T close to TAC*. This is

consistent with mean-field behavior and with previous results

on fully chiral LCs.32,33 Importantly, ec diverges at TAC*, indi-

cating that the observed effect indeed corresponds to coupling

between ~E and the graphene-induced chiral LC domains,

rather than a direct electric field—graphene coupling. We

repeated the experiments at frequencies f¼ 500 Hz and

10 kHz, with no significant differences. It is also important to

point out that the magnitude of ec observed in this experiment

is a few orders of magnitude smaller that of the fully chiral

LCs,23,29 indicating weak chirality at graphene surface.

9OO4 and 5CB are configurationally achiral LCs. But a

chiral conformation in the ground state can be adopted by

rotating left or right22 about the Ph–O bond for 9OO4 and

Ph–Ph bond for 5CB. An equal probability distribution

between left and right twist makes them dynamically race-

mic. In an LCþGP system, the LC director aligns along the

graphene surface,34 employing the p–p electron stacking

between the honeycomb structure of graphene and benzene

rings of the LC to reduce the elastic distortion. When the gra-

phene surface possess strain chirality, the LC can adopt that

surface chirality by shifting its equilibrium between left- and

right-handed chiral conformers to produce a statistically

larger population of chiral conformers of one handedness,

breaking the symmetry between right- and left-handed mo-

lecular conformations: this is also called deracemization. We

observed the electroclinic effect in LC 9OO4 due to this

symmetry breaking in the LC through deracemization.

Parallel graphene sheets present in the LC medium can

induce a twist of the LC director due to the p–p electron

stacking as schematically shown in Figure 1. A twist of the

director can also induce conformational deracemization,28

i.e., chirality. This helical twist (due to chiral induction) can

be experimentally observed in a 90� twist cell.

Due to the chiral induction by graphene flakes, we quali-

tatively observed that the director in the nematic phase of

5CB undergoes a helical twist with a pitch length P, where Pis the distance over which the director rotates by an angle of

2p. We have used LC 5CB for this twist experiment because

of its availability with a room temperature nematic phase. In

this experiment, we used LC cells in which the two surfaces

were arranged to induce a 90� twist of the nematic director

from one substrate to the other. A uniform cell thickness was

maintained by randomly dispersed 6.8 lm spacer particles

during the fabrication of the cells. In the absence of an inher-

ent nematic twist (P!1), there are two possible domains in

the 90� twist cell: one with right-handed twist and the other

with left-handed twist. The spacer particles pin the LC discli-

nations (i.e., the walls between the right-handed and left-

handed domains) in the cell. Since both domains are equally

likely, the disclination lines that run between the cell’s spacer

particles are straight. However, for a chiral liquid crystal with

an intrinsic P, domains having the same sense of surface-

imposed twist grow at the expense of those having the oppo-

site sense of surface-imposed twist, and thus the disclination

lines that connect the spacer particles become bowed. The

equilibrium shape of the disclination line is determined by the

minimization of the total energy, which involves the energy of

the disclination and the twist elastic energy, where it is

assumed that the LC director is anchored strongly at the two

substrates. Thus, a measure of the disclination’s radius of cur-

vature yields the inherent chiral pitch P.24

Figures 4(a)–4(d) show typical polarized micrographs of

pure 5CB at four randomly chosen regions inside a 90� twist

cell. Each micrograph has an area of 1.173� 0.899 mm2. As

expected for a pure achiral LC, the disclination lines between

the equally likely left- and right-handed domains are linear

(P ! 1), minimizing the line tension associated with the

disclination. The dark spots observed in Figure 4 are 6.8 lm

spacer particles, which pin the disclinations. Under the same

conditions, a chiral LC would exhibit curved disclinations in

order to minimize the total energy, and the chiral pitch

length P¼ 2R, where R is the radius of curvature of the

bowed disclination line.24 This method is sensitive to weak

chirality, i.e., large P. Figures 4(e)–4(h) show polarized

micrographs of 5CBþGP at four randomly chosen regions

inside a 90� twist cell. In these four micrographs, clearly

visible curved disclination lines between spacer particles

confirm that the graphene flakes induce chiral twist in the ne-

matic phase. These micrographs also reveal a non-uniform

curvature-distribution throughout the cell. Because the radius

of curvature varies significantly from region to region in the

cell, this behavior suggests a non-uniform density distribu-

tion of graphene flakes in the nematic medium. This curva-

ture variation also suggests that the flakes possess different

chiral strengths due to their size distribution. For the

5CBþGP sample, the radius of curvature R was obtained

from the 1.173� 0.899 mm2 micrographs along several dif-

ferent segments by computer identification and digitization

of the disclination lines. Figure 5 shows the variation of the

chiral pitch length P (¼ 2R) at seven randomly chosenFIG. 3. Inverse of electroclinic coefficient e�1

c vs. T. A linear fit is shown

near TAC* (¼ 62 �C).

114302-3 Basu, Kinnamon, and Garvey J. Appl. Phys. 118, 114302 (2015)

[This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to ] IP:

73.163.157.169 On: Wed, 16 Sep 2015 14:35:50

regions (i.e. seven 1.173� 0.899 mm2 micrographs) in the

5CBþGP cell. Regions 1 to 4 in Figure 5 correspond to

micrographs (e) to (h) in Figure 4, respectively. With a wide

variation, the average P is found to be 2.9 mm. This large Pindicates that the graphene-induced chirality is weak——

this is consistent with weak electroclinic signal for the 9OO4

þGP sample.

III. SUMMARY

Based on transmission electron microscopy, electron

energy loss spectroscopy, Raman spectroscopy, and electron

diffraction35 data from the manufacturer, there is no sugges-

tion of chiral impurities in the graphene sample (i.e., 99.99%

carbon content). The appearance of an electroclinic effect for

the graphene-doped 9OO4 LC and its absence for pure

9OO4 is, therefore, an unambiguous signature of induced

chirality in an achiral LC, arising from the graphene flakes.

On the other hand, the presence of bowed disclinations for

the GP-doped-5CB LC and linear disclinations for pure 5CB

is a distinct characteristic of induced chirality and propaga-

tion of twist through nematic correlation in an achiral ne-

matic LC. These two independent experiments coherently

suggest that the chirality of graphene can be detected using

FIG. 4. Micrographs (1.173

� 0.899 mm2) showing disclination

lines in two 90� twist cells under the

cross polarized microscope; (a)–(d)

pure 5CB showing straight disclination

lines at four different places in the

cell; (e)–(h) 5CBþGP showing bowed

disclination lines at four different pla-

ces in the cell. Black dots are 6.8 lm

spacer particles.

114302-4 Basu, Kinnamon, and Garvey J. Appl. Phys. 118, 114302 (2015)

[This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to ] IP:

73.163.157.169 On: Wed, 16 Sep 2015 14:35:50

LC platforms. These results suggest that the graphene sample

contains a small quantity of an enantiomeric excess.

However, the origin of this enantiomeric excess of graphene

is still unknown. The chiral induction in the LC may occur in

two different ways in this case. The strain chirality on gra-

phene surface can break the symmetry between the right-

and left-handed molecular conformations in the LC. Another

way, the dispersed graphene flakes self-align in the LC with

their flat surface parallel to the LC director.36 Therefore,

when they are parallel to each other, they can also induce a

macroscopic twist on the LC director, inducing deracemiza-

tion in the LC. In addition to the existing chirality detection

techniques of graphene, such as circular dichroism1 and

Raman spectroscopy,10 this new technique with the LCs will

not only make graphene technologically useful but will also

lead directly to the central role played by LC-influenced

organizational themes in the development of nanoscale self-

assembled systems, such as the propagation of chirality from

the graphene surface to the LC.

ACKNOWLEDGMENTS

This work was supported by the Office of Naval Research

(Award No. N0001415WX01534) and the investment grant at

the U.S. Naval Academy.

1T. Jiang, D. Emerson, K. Twarowski, D. Finkenstadt, and J. Therrien,

Phys. Rev. B 82, 235430 (2010).

2J. Cao, H.-J. Yin, and R. Song, Front. Mater. Sci. 7, 83 (2013).3X. L. Li, X. R. Wang, L. Zhang, S. W. Lee, and H. J. Dai, Science 319,

1229 (2008).4M. Y. Han, B. Ozyilmaz, Y. B. Zhang, and P. Kim, Phys. Rev. Lett. 98,

206805 (2007).5Y. W. Son, M. L. Cohen, and S. G. Louie, Phys. Rev. Lett. 97, 216803

(2006).6C. Tao, L. Jiao, O. V. Yazyev, Y.-C. Chen, J. Feng, X. Zhang, R. B.

Capaz, J. M. Tour, A. Zettl, S. G. Louie, H. Dai, and M. F. Crommie,

Nature Phys. 7, 616 (2011).7L. Yang, M. L. Cohen, and S. G. Louie, Nano Lett. 7, 3112 (2007).8A. G. Moghaddam and M. Zareyan, Appl. Phys. A: Mater. Sci. Process.

89, 579 (2007).9Y. W. Son, M. L. Cohen, and S. G. Louie, Nature (London) 444, 347

(2006).10Y. You, Z. Ni, T. Yu, and Z. Shen, Appl. Phys. Lett. 93, 163112 (2008).11M. Begliarbekov, K.-I. Sasaki, O. Sul, E.-H. Yang, and S. Strauf, Nano

Lett. 11, 4874 (2011).12R. Basu, Appl. Phys. Lett. 105, 112905 (2014).13R. Basu, A. Garvey, and D. Kinnemon, J. Appl. Phys. 117, 074301 (2015).14A. Malik, A. Choudhary, P. Silotia, A. M. Biradar, V. K. Singh, and N.

Kumar, J. Appl. Phys. 108, 124110 (2010).15T. M. Alam and C. J. Pearce, Chem. Phys. Lett. 592, 7 (2014).16A. B. Shivanandareddy, S. Krishnamurthy, V. Lakshminarayanan, and S.

Kumar, Chem. Commun. 50, 710 (2014).17P. Blake, P. D. Brimicombe, R. R. Nair, T. J. Booth, D. Jiang, F. Schedin,

L. A. Ponomarenko, S. V. Morozov, H. F. Gleeson, E. W. Hill, A. K.

Geim, and K. S. Novoselov, Nano Lett. 8, 1704 (2008).18Y. U. Junga, K. W. Parka, S. T. Hura, S. W. Choia, and S. J. Kanga, Liq.

Cryst. 41, 101 (2014).19R. Basu, K. Boccuzzi, S. Ferjani, and C. Rosenblatt, Appl. Phys. Lett. 97,

121908 (2010).20R. Basu, R. G. Petschek, and C. Rosenblatt, Phys. Rev. E 83, 041707

(2011).21R. Basu, C.-L. Chen, and C. Rosenblatt, J. Appl. Phys. 109, 083518

(2011).22R. Basu, C. Rosenblatt, and R. Lemieux, Liq. Cryst. 39, 199 (2012).23S. Garoff and R. B. Meyer, Phys. Rev. Lett. 38, 848 (1977).24E. P. Raynes, Liq. Cryst. 33, 1215 (2006).25S. Ferjani, Y. Choi, J. Pendery, R. G. Petschek, and C. Rosenblatt, Phys.

Rev. Lett. 104, 257801 (2010).26M. Kimura, Y. Yamada, H. Ishihara, and T. Akahane, Mol. Cryst. Liq.

Cryst. Sci. Technol., Sec. A 302, 199 (1997).27J.-Z. Xue and N. A. Clark, Phys. Rev. Lett. 64, 307 (1990).28R. Basu, J. Pendery, R. G. Petschek, R. P. Lemieux, and C. Rosenblatt,

Phys. Rev. Lett. 107, 237804 (2011).29G. Andersson, I. Dahl, P. Keller, W. Kuczynski, S. T. Lagerwall, K.

Skarp, and B. Stebler, Appl. Phys. Lett. 51, 640 (1987).30V. Jayalakshmi, T. Wood, R. Basu, J. Du, T. Blackburn, C. Rosenblatt, C.

M. Crudden, and R. P. Lemieux, J. Mater. Chem. 22, 15255 (2012).31Z. Li, R. G. Petschek, and C. Rosenblatt, Phys. Rev. Lett. 62, 796 (1989).32R. Qiu, J. T. Ho, and S. K. Hark, Phys. Rev. A 38, 1653 (1988).33Z. Li and C. Rosenblatt, Phys. Rev. A 39, 1594 (1989).34D. W. Kim, Y. H. Kim, H. S. Jeong, and H.-T. Jung, Nature Nanotechnol.

7, 29 (2012).35A. Dato, V. Radmilovic, Z. Lee, J. Phillips, and M. Frenklach, Nano Lett.

8, 2012 (2008).36R. Basu, D. Kinnamon, and A. Garvey, Appl. Phys. Lett. 106, 201909

(2015).

FIG. 5. The distribution of pitch length P for 5CBþGP at seven randomly

chosen regions in the 90� twist cell. The dotted line shows the average P.

114302-5 Basu, Kinnamon, and Garvey J. Appl. Phys. 118, 114302 (2015)

[This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to ] IP:

73.163.157.169 On: Wed, 16 Sep 2015 14:35:50