Carbon nanotubes length optimization for preparation of improved transparent and conducting thin film substrates

Original ArticleCarbon nanotubes length optimization for preparation of improved Department of Physics, Shahid Chamran University of Ahvaz, Ahvaz, Islamic Republic of Iran a r t i c l e

Original Article

Carbon nanotubes length optimization for preparation of improved

Department of Physics, Shahid Chamran University of Ahvaz, Ahvaz, Islamic Republic of Iran

a r t i c l e i n f o

Article history:

Received 4 November 2016

Received in revised form

18 February 2017

Accepted 19 February 2017

Available online 27 February 2017

Keywords:

Transparent and conductive films

Multi-walled carbon nanotubes

Spin coating

Figure of merit

Substrate

a b s t r a c t

Transparent and conductive thin films of multiwalled carbon nanotubes (MWCNTs) with different lengths were prepared on glass substrates by the spin coating method In order to reduce the MWCNTs length, they were functionalized The initial length of MWCNTs (10e15mm) was reduced to 1200, 205 and 168 nm after 30, 60 and 120 min refluxing time, respectively After post annealing at 285
C for 24 h, the electrical and optical properties were greatly improved for functionalized MWCNT thinfilms They strongly depend on the length of CNTs The optical transmittance of thefilm prepared using 30 min reflux CNTs was 2.6% and 6.6% higher than that of the 60 min and 120 min refluxed samples respectively The sheet resistance of thisfilm showed reductions of 45% and 80% as well The film also exhibited the least roughness The percolativefigure of merit, which is proportional to the transparency and dispro-portional to the sheet resistance, was found to be higher for the sample with 30 min refluxed MWCNTs

© 2017 The Authors Publishing services by Elsevier B.V on behalf of Vietnam National University, Hanoi This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/)

1 Introduction

Transparent and conductive thinfilms have attracted

consider-able interest due to their importance in the fundamental researches

and potential industrial applications in optoelectronic devices,

transparentfilms, automobile glasses and smart windows Indium

tin oxide (ITO) is a superior material which is used to fabricate the

transparent conductingfilms[1e3] For these materials, the main

disadvantage is related to its applicability toflexible substrates The

small strains of these materials cause a reduction in their electrical

conductivity and so do in their performance as a good conducting

film Currently, carbon nanotubes are the materials of

ever-increasing interest due to their excellent electronic, physical and

chemical properties and are of major research interest for their

outstanding behaviours in practical applications[4e8] It has been

shown that the single wall carbon nanotubes' performance is

comparable to ITO for many applications such as solar cells, smart

windows and sensors[9,10] The high electrical conductivity and

mechanical strength of carbon nanotubes (CNTs) make them a good

replacing candidate for the ITO materials to make transparent

conducting films The CNTs transparent conducting films can be prepared onflexible substrates so they may have many applications

in different types of electronic, optoelectronic, solar cell and sensor systems[10] Different conductivities have been reported for CNTs because several factors can affect the conductivity of CNTs such as sample purity, metallic to semiconducting volume ratio and doping level of the semiconducting CNTs[10]

In order to investigate the relation between the optical and electrical properties of CNTs' films, some parameters need to be introduced It has been shown that the below relation is held be-tween transmittance (T), optical conductivity (sop) andfilm thick-ness (t)[11]:

T¼

1þZ0

2sopt

2

(1)

where the Z0is the free space impedance (377U) This relation can

be converted to a relation between T and sheet resistance (Rs) as

T¼

1þ Z0

2Rs

sop sDC:B

2

(2)

where sDC.B is the bulk DC conductivity For a thinfilm, DC conductivity is thickness dependent and is proportional to tn [12e15], where t is the film thickness and n is the percolation

* Corresponding author Fax: þ98 6133331040.

E-mail addresses: Farbod_m@scu.ac.ir (M Farbod), amirzilaie@gmail.com

(A Zilaie), I.Kazeminezhad@scu.ac.ir (I Kazeminezhad).

Peer review under responsibility of Vietnam National University, Hanoi.

Contents lists available atScienceDirect Journal of Science: Advanced Materials and Devices

j o u r n a l h o m e p a g e : w w w e l s e v i e r c o m / l o c a t e / j s a m d

http://dx.doi.org/10.1016/j.jsamd.2017.02.005

2468-2179/© 2017 The Authors Publishing services by Elsevier B.V on behalf of Vietnam National University, Hanoi This is an open access article under the CC BY license

Journal of Science: Advanced Materials and Devices 2 (2017) 99e104

exponent Using a simplified model, De et al have found another

relation between T and Rsof transparent conductors in percolative

regime at which the conductivity is thickness dependent as[15]:

T¼ 1þQ1

Z0

Rs

1

þn!2

(3)

here,Pis a dimensionless parameter called the percolativefigure

of merit and its higher values mean tminhigher T and lower Rs

In order to rate the performance of differentfilms, we applied

these equations to thefilms prepared by different lengths CNTs and

found different behaviours that will be presented and discussed in

this paper We believe that the length is an important factor which

can have a significant impact on the conductivity and transparency

of CNTs thinfilms In particular, in the present study, MWCNTs thin

films with different CNTs lengths were fabricated by spin coating

method without using any surfactant The sheet resistance and

optical transmittance of thefilms were measured and the findings

were interpreted based on the CNTs lengths

2 Experimental

2.1 Materials

CVD synthesized pristine MWCNTs with the length of

10e15 mm and diameters of less than 10 nm were purchased

(Shenzhen Co., China) Commercial round glasses with 15e19 mm

in diameter were also used as substrate Purification and

func-tionalization of the MWCNTs were carried out into a mixture of

concentrated sulphuric and nitric acids (95% H2SO4, 65% HNO3;

3:1) Details of acid treatment have been described elsewhere[16] The chemical functionalization of CNTs is a technique to improve their dispersibility and homogeneity in organic solvents and water

2.2 Surface modification of the glass substrate

In order to have better adhesion between carbon nanotubes and the glass substrates, 3-aminopropyltriehtoxysilane (APTES) solu-tion was used to modify the substrate's surface The glass substrate was cleaned by DI water, acetone and isopropanol alcohol respec-tively in an ultrasonic bath for 15 min then dried in an oven at

100
C for 30 min The dried substrate then was immersed into a prepared aqueous solution including 0.5 ml of APTES mixed with

47 ml deionized water for 24 h[17,18] It was found that without using APTES, the adhesion between glass surface and CNTs was very weak and the CNTs film could easily peeled-off from the substrate

2.3 Film fabrication method Four kinds of acid-treated MWCNTs which were refluxed for different times of 15, 30, 60 and 120 min were used to make a suspension One milligram of such MWCNTs was dispersed into

1 ml of pure ethanol by an ultrasonic bath (42 kHz, 100 W) for

15 min The suspension then centrifuged for a few minutes to remove the possible existing particles and large bundles in the suspension The spin coating method was employed as a fast technique to prepare thefilms A final speed of 4900 rpm for 25 s was chosen Thefilms then were annealed at 285
C for 24 h.

films prepared using different refluxed times of MWCNTs and a typical 135 nm thin film on glass substrate.

2.4 Characterization

Afield emission scanning electron microscope (FESEM: MIRA,

TESCAN- Czech Republic) was used to study the surface

morphology and thickness of MWCNTsfilms To measure the film's

thickness, the glass substrates were tilted for a better view for

imaging Sheet resistance of thefilms was measured by 4-point

probe technique at room temperature The optical transmittance

was recorded be means of a UVeVis spectrophotometer (Cintra 101,

GBCe Australia) The roughness of the films was measured using

an atomic force microscope (SPM: DME, 95-50E- Denmark)

3 Results and discussion

The SEM images offilms prepared using 30, 60 and 120 min

refluxing time and a typical 135 nm thickness on glass substrate are

shown inFig 1 As can be observed, by increasing the refluxing

time, the lengths of CNTs become smaller During the

functionali-zation, the carbon nanotubes break apart and their length

decreased which was depending on the refluxing time It was found

from a“computer software measurement” that the initial length of

CNTs (10e15mm) is reduced to 1200, 205, 168 nm after 30, 60 and

120 min refluxing time respectively

A circuit set up to light an LED with a battery, using 30 min

refluxed CNTs films with various thicknesses in series with a

bat-tery and LED is presented inFig 2 The optical transmittance of the

films prepared using different refluxed CNTs with different

thick-nesses is illustrated inFig 3 As can be observed, the transmittance

increases with decreasing the thickness for all thefilms

In order to describe thefigures' differences, the transmittances

of thefilms at 550 nm were plotted versus their thickness and are

shown inFig 4 It is clear seen from thefigure that by increasing the

films' thicknesses, the transmittance decreases The decrease rate is

nearly the same for the films prepared using CNTs refluxed for

higher than 30 min A faster decrease, however, is observed for

films prepared by 15 min refluxed CNTs For all thicknesses, the

transmittance is almost highest for 30 min refluxed CNTs films

Indeed, the optical transmittance of the 30 min refluxed films was

2.6% and 6.6% higher than that of for 60 min and 120 min ones This

means that besides the thickness, the length of CNTs can affect the

transmittance of thefilms Due to the fact that the 15 min refluxed

CNTsfilms exhibited unsuitable behaviours, their data will not be

presented here after

The sheet resistances of the films were calculated using IeV measurements.Fig 5shows a typical IeV graph of different films with the same thickness and different CNTs lengths The sheet resistance deduced using such measurements are plotted versus their thickness and shown inFig 6 As can be seen from thefigure, the sheet resistance reduces with increasing the thickness expo-nentially for the all samples The sheet resistance value of 30 min

reflux CNTs films showed 45% and 80% reduction compared to the

60 min and 120 min refluxed films, respectively Also the sheet Fig 2 Transparency of a typical film and a circuit set up to light an LED, using the film

Fig 3 Optical transmittance of thin films prepared using different length function-alized MWCNTs and different thicknesses The higher refluxing time means the length

of CNTs are shorter.

M Farbod et al / Journal of Science: Advanced Materials and Devices 2 (2017) 99e104 101

resistance shows no significant thickness dependence when the

thickness is higher than 300 nm In addition, the sheet resistance of

such film showed 45% and 80% reduction as well Although the

samples are exposed to the difference in refluxing time, the

dif-ference in optical and electrical properties can not be connected to

the different density of defects Based on our previous work[16]on

Raman study of functionalized CNTs, it was found that after 1 h

refluxing, the ratio of ID/IGwas kept constant This indicated that the difference in above mentioned properties is independent on the defect levels, but can be affected by the length of CNTs

As already mentioned in Section2.3, thefilms were annealed at

285
C after being spin-coated We found that the post heat treatment has a remarkable influence on the reduction of sheet resistance and the improvement of thefilms' optical transmittance

By post heat treatment the sheet resistance was decreased at least one order of magnitude It seems that before annealing, the po-tential barrier between the CNTs junctions is high and the free carriers are impeded so the electrical conductivity is poor By post annealing the contacts and fusion between the CNTs is improved resulting in the reduction of the sheet resistance

Fig 7shows the plot of the optical transmittance at 550 nm versus sheet resistance of the films prepared using CNTs with different lengths As it was expected, thefilms with higher trans-parency typically showed higher sheet resistance It is desirable to havefilms with high transparency and low sheet resistance From theFig 7it is clear that the transparency of thefilms with the same sheet resistance depends on the length of CNTs and the films which were prepared using 30 min refluxed CNTs show higher transparency It seems that the settlement of CNTs on the substrate during the film coating is arranged in a manner that

Fig 4 Transmittance versus thickness for the films prepared using different refluxed

time CNTs.

Fig 5 Typical IeV graph of films prepared using different length CNTs and the same

thickness of about 240 nm The higher slope means lower resistance.

Fig 6 Sheet resistance versus thickness for films prepared using differently refluxed

CNTs.

Fig 7 Optical transmittance at 550 nm versus sheet resistance for the films prepared using different lengths CNTs.

Fig 8 Average roughness versus the thickness for films prepared using differently refluxed CNTs.

shorter CNTs allow for less light transmission Such observation is

in agreement with thefilm roughness measurements

Fig 8shows the average roughness versus thefilm's thickness

which were measured using AFM images One can observe that by

increasing thefilms' thickness their roughness decrease As it is

clear from the figure, the roughness for the films which were

prepared using 30 min refluxed CNTs is lower than that of the

others

Further analysis of CNTsfilms was carried out by studying the

percolation behaviour of the network By drawing a vertical line at a

certain thickness inFig 6one can observe that at the same

thick-ness, the sheet resistance is lesser for thefilms which were

pre-pared using 30 min refluxed CNTs As mentioned above, by

increasing the refluxing time from 30 to 120 min the initial length

of CNTs reduced to 1200, 205 and 168 nm Therefore, for thefilms

with the same thicknesses, the sheet resistance is higher when the

CNTs lengths are shorter According to Hecht et al.[19]for a CNTs

network two kinds of resistance sources are existed One is the

resistance along the MWCNT itself (RNT) and the other is due to the

CNTseCNTs junction (Rjnt) When RNT>> Rjnt, the sheet resistance

or conductivity should be independent of the length of CNTs but if

Rjnt>> RNTthe sheet resistance of the network should be

depen-dent on the number of CNTseCNTs junctions So, using shorter CNTs

infilm fabrication means that the number of CNTseCNTs junctions

increased, which leads the sheet resistance enhancement

It has been reported that for a network of Ag nanowires with a

given length (L), the critical number of nanowires (Nc) required for

percolation is given by the below equation[18]:

NcL2¼ 5:71

If we apply such an equation to CNTs thin films, one can

conclude that an electrical percolation path can be achieved by less

number of CNTs but with longer length If more CNTs with shorter

length are used, more connections will be formed resulting in

higher sheet resistance and lower transmittance Fig 9 shows

schematically that a percolation path can be achieved by a thinner

layer but longer carbon nanotubes

Based on our data,sopof the films and the sop

sDC:B ratio were calculated using the Eqs (1) and (2) respectively In order to

calculate the films percolative figure of merite (P) and n the

percolation exponent, the Eq (3) was employed They were

deduced using the logelog plot of the (T0.5 1) versus Rs The

calculated parameters are listed inTable 1

The results show that the percolativefigure of merit is higher for

thefilms with longer CNTs This means utilizing 30 min refluxed

CNTs for transparent thinfilm fabrication can give better results So

we suggest being careful in using refluxed CNTs for fabrication of

transparent conductingfilms

4 Conclusion Carbon nanotube thinfilms with different thickness and CNTs lengths were fabricated Their sheet resistance and optical trans-mittance were investigated It was observed that the sheet resistance and optical transmittance of MWCNTs films were extremely dependent on the CNTs average length The optical transmittance of the films prepared using 30 min reflux CNTs was 2.6% and 6.6% higher than that of for 60 min and 120 min refluxed films with the same thickness, respectively Also the sheet resistance of 30 min

reflux CNTs films showed 45% and 80% reduction compared to the

60 min and 120 min refluxed films, respectively Such films showed the least roughness among the other samples The percolativefigure

of merit was 4.5, 3.3 and 2.5 for thefilms prepared using 30, 60 and

120 min refluxed CNTs respectively It proposes a possibility to achieve percolation conducting paths by longer length carbon nanotubes at lower thickness and therefore higher transmittance Acknowledgments

The authors acknowledge Shahid-Chamran University of Ahvaz for thefinancial support of this work and also Khuzestan Water and Power Authority for their helps to use their laboratory

References [1] H.E Unalan, G Fanchini, A Kanwal, A.D Pasquier, C Manish, Design criteria for transparent single-wall carbon nanotube thin film transistors, Nano Lett 6 (2006) 667e682

[2] T Minami, Present status of transparent conducting oxide thin-film devel-opment for Indium-Tin-Oxide (ITO) substitutes, Thin Solid Films 516 (2008) 5822e5828

[3] X.W Sun, H.C Huang, H.S Kwok, On the initial growth of indium tin oxide on glass, Appl Phys Lett 68 (1996) 2663e2665

[4] V.H.R Souza, S Husmann, E.G.C Neiva, F.S Lisboa, L.C Lopes, R.V Salvatierra, A.J.G Zarbin, Flexible, transparent and thin films of carbon nanomaterials as electrodes for electrochemical applications, Electrochim Acta 197 (2016) 200e220

[5] P Liu, Y.F Tan, D.C.M Hu, D Jewell, H.M Duong, Multi-property enhancement

of aligned carbon nanotube thin films from floating catalyst method, Mater Des 108 (2016) 754e760

[6] T Abendroth, H Althues, G M€ader, P H€artel, S Kaskel, E Beyer, Selective absorption of carbon nanotube thin films for solar energy applications, Sol Energy Mater Sol C 143 (2015) 553e556

[7] S Mallakpour, E Khadem, Carbon nanotubeemetal oxide nanocomposites: fabrication, properties and applications, Chem Eng J 302 (2016) 344e367 [8] X Zheng, J.S Kim, C.W Park, Multiwall carbon nanotube/copper porous coating for heat transfer application, Surf Eng 31 (2015) 723e732 [9] H Hanaei, M.K Assadi, R Saidur, Highly efficient antireflective and self-cleaning coatings that incorporate carbon nanotubes (CNTs) into solar cells:

a review, Renew Sustain Energy Rev 59 (2016) 620e635 [10] M Kaempgen, S Roth, Transparent and flexible carbon nanotube/polyaniline

pH sensors, J Electroanal Chem 586 (2006) 72e76 [11] M Dressel, G Grüner, Electrodynamics of Solids: Optical Properties of Elec-trons in Matter, Cambridge University press, Cambridge, 2002

[12] B Dan, G.C Irvin, M Pasquali, Continuous and scalable fabrication of trans-parent conducting carbon nanotube films, ACS Nano 3 (2009) 835e843 [13] L Hu, D.S Hecht, G Gruner, Percolation in transparent and conducting carbon nanotube networks, Nano Lett 4 (2004) 2513e2517

[14] D.S Stauffer, A Aharony, Introduction to Percolation Theory, Taylor, Francis, London, 1994

[15] S De, P.J King, P.E Lyons, U Khan, J.N Coleman, Size effects and the problem with percolation in nanostructured transparent conductors, ACS Nano 4 (2010) 7064e7072

[16] M Farbod, S Khajehpour-Tadavani, A Kiasat, Surface oxidation and effect of electric field on dispersion and colloids stability of multiwalled carbon nanotubes, Colloids Surf A Physicochem Eng Asp 384 (2011) 685e690

Fig 9 A percolation path from A to B can be achieved by a thinner layer but longer length carbon nanotubes.

Table 1

Values ofsDC.B /sOp ,Pand n found from fitting in curves of (T 0.5 1) versus R s for

different films.

Refluxed time (min) sDC.B /sop sop (S/m) n Q

M Farbod et al / Journal of Science: Advanced Materials and Devices 2 (2017) 99e104 103

[17] A Gomathi, S.R.C Vivekchand, A Govindaraj, C.N.R Rao, Chemically bonded

ceramic oxide coatings on carbon nanotubes and inorganic nanowires, Adv.

Mater 17 (2005) 2757e2761

[18] T.H Joong, Y.K Sun, S.W Jong, L Geon-Woong, Transparent, conductive and

super hydrophobic films from stabilized carbon nanotube/silane sol mixture

solution, Adv Mater 20 (2008) 3724e3727

[19] D Hecht, L Hu, G Grüner, Conductivity scaling with bundle length and diameter in single walled carbon nanotube networks, Appl Phys Lett 89 (2006) 113112e113113