Supporting information

Damage mitigation in roll-to-roll transfer

of CVD-graphene to flexible substrates

Bongkyun Jang1,2, Chang-Hyun Kim1,3, Seung Tae Choi4, Kyung-Shik Kim1, Kwang-Seop

Kim1,3, Hak-Joo Lee1,5, Seung-Min Cho6, Jong-Hyun Ahn7, Jae-Hyun Kim1,3,∗

1Department of Nano Mechanics, Nano-Convergence Mechanical Systems Research

Division, Korea Institute of Machinery and Materials (KIMM), Daejeon 34103, Republic of

Korea

2Material Science Laboratory, Graduate School of Engineering, Department of Mechanical

Engineering and Science, Kyoto University, Kyoto 615-8540, Japan

3Nano Mechatronics, Korea University of Science and Technology (UST), Daejeon 34113,

Republic of Korea

4School of Mechanical Engineering, Chung-Ang University, Seoul 06974, Republic of Korea

5Center for Advanced Meta-Materials (CAMM), Daejeon 34103, Republic of Korea

6New Business Division, Hanhwa Techwin R&D Center, Gyeonggi-do 13488, Republic of

Korea

7School of Electrical and Electronic Engineering, Yonsei University, Seoul 03722, Republic

of Korea

*Corresponding Author : jaehkim@kimm.re.kr

S1 Fabrication process of graphene transparent conductive films

Figure S1. Transfer process of CVD graphene synthesized on copper foil for a fabrication of

flexible transparent electrode.

S2. Temperature dependence on TRT-based graphene transfer process

Figure S2. (a) Temperature dependence of sheet resistance of transferred graphene onto PET

substrate. Inset figures are SEM images of TRT surface after transfer. (b) TRT residues on

graphene transferred on PET after transferring with the transfer temperature of 130 °C, and

(c) its corresponding surface morphologies of TRT film after heating.

S3 Control of nip force in roll to roll transfer machine

Figure S3. Nip forces, Fz1 and Fz2, measured by load sensors equipped in the both end of the

nip rolls (a) without control, (b) with control of nip force.

S4 Integrated control system for continuous roll to roll transfer of graphene

Figure S4. Force diagram of the roll transfer system with the two rollers and tension

measurement module between them.

Figure S4 shows a force diagram of the roll transfer system integrated with the force control

module. Films are continuously fed from the left side and pass through the first nip rolls,

tension control module, and the second nip rolls, step by step. Each roll pair composed of a

master roll (upper side) and a slave roll (bottom side). The axis of the master roll is fixed, but

the axis of the slave roll is connected with the translation stage in vertical direction to control

nip force uniformly by the load cells to measure the force of normal directions. The load cells

to measure the force of lateral directions, equipped in the axis of the slave roll, can detect

shear forces applied to the films. For convenience, the index of the first master roll, the first

slave roll, the second master roll, and the second slave roll is referred as M1, S1, M2, and S2,

respectively. When the system is stably controlled, the linear velocity is expressed as, v = R,

where R is a radius of rolls, and is an angular velocity of rolls. Here, it is possible to define

average linear velocity and relative linear velocity for each nip roll like,

v1,2=vM 1,2+vS 1,2

2= R

2 ( ωM 1,2+ωS1,2 ) ,∆ v1,2=vM 1,2−vS1,2=R ( ωM 1,2−ωS1,2 ) .

To establish a stable operation, tension is applied films fed to the system. Since nip rolls

isolate tensions, it is possible to consider the three tension values in the system, T1, T2, and T3.

T2 is directly measured with a tension load cell installed in the tension control module with

the relation of FT=2T 2sin θ. If the shear force applied between nip rolls is referred as FM1,

and FM2, the lateral forces are measured with x-axis load cell as,

F x 1=FM 1+T2−T1 ,F x 2=FM 2+T3−T2 .From this equation, the lateral forces are related with

tensions of the adjacent films with nip rolls, and the shear forces generated from nip rolls.

Generally, the shear force is dependent on the normal force and rotation velocity of the nip

rolls [S1]. So, the shear force also expressed as,

FM 1,2=μ F z1,2[1−(1−κ1,2Δv1,2

v1,2 )2] ,where

κ1,2=πGal

4 (1−ν ) μ F z1,2

.

Here, , G, , a, and l are frication coefficient, shear modulus, Poison’s ratio, and two sides

of rectangular contact area, respectively. This equation means that the shear forces of nip rolls

is correlated with the normal forces since friction force is large when the normal contact

pressure increases. Also, the difference of relative velocity causes an increase of the shear

force. So, the synchronization of nip rolls and the small normal force can minimize the shear

force. However, the current roll transfer system displayed in figure S4 includes two nip roll

pairs. Therefore, the integrated control system should be considered for a minimization of the

shear force.

Figure S5. Force measured by each load cells and linear speed of the transfer machine. For

initial disturbances, the system is stably controlled after proceeding of 300 mm.

S5 High speed R2R transfer of graphene

Figure S6. High speed transfer of graphene with tuned transfer parameters. (a) Control of nip

forces, lateral forces, and tension of the integrated system with a translation speed of 2000

mm/min. (b) Sheet resistance map of the transferred graphene at 2000 mm/min. (c)

Photograph of roll transfer process and (d) the transferred graphene on PET with a width of

500 mm. Higher speed transfer of graphene necessitates a fine tuning of transfer temperature,

nip force, lateral force and tension.

S6 Effects of synchronization between nip rolls

Figure S7. Rotational synchronization of nip rolls. (a) Schematics of synchronization test. (b)

Sheet resistance changes of the example case with synchronized setting and different velocity

setting of upper and lower roll. (c) Parallel wrinkles of graphene transferred on PET, which

is observed in extreme case of asynchronous rotation velocity. (d) Warpage of

TRT/graphene/PET structure laminated without synchronization of rotational velocity.

S7 Characterization of surface morphology

Figure S8. SEM images of (a) graphene on Cu foil and (b) graphene on the carrier film

(TRT).

Figure S9. SEM image and filtered images of graphene transferred on f the flexible substrate

(PET) for identify cracks and defects clearly.

S8 Mechanical properties for finite element analysis

Table S1. Young’s modulus and Poisson’s ratio of each material used in the analysis

Young’s modulus Poisson’s ratio

Graphene 1.0 TPa 0.165

PET 2.0 GPa 0.4

Rubber coating 6.06 MPa 0.5

TRT adhesive 20 kPa 0.5

Mechanical properties of graphene is estimated by density function theory simulation and is

proved by experiments [2, S2]. Young’s modulus of TRT adhesive is assumed as a modulus

measured by tensile test for conventional adhesive layer, which is similar types of TRT

adhesive [S3]. Also, the properties of PET and rubber coating of the rollers are measured by

uniaxial tensile test and hardness test, respectively.

S9 Deformation of adhesive layer calculated by finite element analysis

Figure S10. Strain distribution of an adhesive layer of a carrier film used in graphene transfer,

with respect to displacements. (a) Wavy-structured adhesive layer with graphene layer, and

(b) wavy structured adhesive layer without graphene layer.

To investigate deformations of adhesive in a carrier film during the transfer process, finite

element analysis is performed using commercial finite element method code, ABAQUS. The

adhesive layer with 60 μm thickness and wavy surface is compressed on the flat and rigid

substrate.

Figure S10 illustrates maximum principal strain distribution of the adhesive layer. The

adhesive layer without graphene layer deform and the contact region of it increase under

compression as shown Figure S10(b). Before fully contact state, δ = 180 nm, the crest start to

contact to the substrate and the adhesive largely deforms near it. After entire surface of the

adhesive fully contact with the substrate at δ = 182 nm, strain near the surface is under 0.1. In

addition, the wavy surface of the adhesive establish a conformal contact with the flat surface.

This result can explain the reason why the roughness of bare TRT increase after lamination

with graphene on Cu foil, and etching of Cu foil. When it comes to contact between rough

and rigid graphene on Cu, and relatively flat TRT, the adhesive of TRT largely deforms

viscoplastically and conform the surface of the graphene. After etching of Cu, the surface of

graphene on TRT resembles the surface of graphene on Cu foil.

However, when thin and stiff graphene layer is covered on the wavy surface of the adhesive

layer, the adhesive layer largely deform in local area compared to the surface without

graphene, as shown in figure S10(a). This is because the buckling of graphene layer on the

adhesive under the compression. Because of the mismatch of the modulus between the

adhesive and the graphene, relatively stiff graphene layer is buckled. Even though, high

normal stress is applied on the adhesive, the surface cannot fully contact with the flat

substrate. This buckling mechanism can explain the damage induced by normal stress in the

wavy structure of the adhesive.

S10 Lamination and transfer with small nip forces

During the lamination and transfer process in the R2R transfer of graphene, small nip forces

cause failures such as air bubbles and wrinkles. Figure S11 shows the photographs of those

failures obtained from the lamination process between a Cu foil and a TRT with small nip

forces. The air bubbles and wrinkles are mainly originated from the incomplete contact

between graphene and the target film. The incomplete contact is inevitable for small nip force

because of roundness error of the rollers and uneven thicknesses of TRT and the target film.

To make complete contact, it is necessary to increase the nip force. From the experimental

data, we found that the nip force per unit width should be larger than 3.27 N/mm for the

present case.

Figure S11. Air bubbles and wrinkles generated from the laminated films by R2R process

with small nip forces.

Reference

[S1] Stolarski, T. A. and Tobe, S. Rolling Contacts. (Professional Engineering Publishing

Limited, London, UK, 2000).

[S2] Zhou L, Wang Y, Cao G. 2013 Elastic properties of monolayer graphene with different

chiralities. J. Phys.: Condens. Matter. 25 125302

[S3] Murata A, Oshima T, Arimitsu Y and Kiuchi K 2010 Heat-peelable pressure-sensitive

adhesive sheet US Patent 7,718,257