Synthesis of ultra-thin copper nanowires using tris...

1

Supporting Information for:

Synthesis of ultra-thin copper nanowires using

tris(trimethylsilyl)silane for high-performance and low-haze

transparent conductors

Fan Cui, 1,2,3 Yi Yu,1,3 Letian Dou, 1,2,3 Jianwei Sun, 1,3 Qin Yang, 1 Christian

Schildknecht,2 Kerstin Schierle-Arndt2 and Peidong Yang1,2,3,4,5 *

1Department of Chemistry, University of California, Berkeley, CA 94720, USA. 2California

Research Alliance (CARA), BASF Corporation, Berkeley, CA 94720, USA. 3Materials Sciences

Division, Lawrence Berkeley National Laboratory, Berkeley, CA 94720, USA. 4Kavli Energy

NanoSciences Institute at the University of California, Berkeley and the Lawrence Berkeley

National Laboratory, Berkeley, California 94720, USA. 5Department of Materials Science and

Engineering, University of California at Berkeley, Berkeley, California 94720, USA.

*Correspondence and requests for materials should be addressed to P. Yang (email:

[email protected]).

2

Materials and Methods

Regents:

Tris(trimethylsilyl)silane (TTMSS, 97%), copper (II) chloride dihydrate (CuCl2·2H2O,

99.999%), oleylamine (70%), oleic acid (90%), and nitrocellulose filter membranes (25

mm diameter, 220 nm pore size), were purchased from Sigma-Aldrich and used as

received. Unless otherwise stated, all of other chemicals were purchased from Sigma-

Aldrich and used as received.

Experimental details:

Synthesis of copper nanowires:

In a typical reaction, 85 mg of CuCl2·2H2O (0.5 mmol) and 5 g of oleylamine were

mixed in a reaction vessel and then sonicated at room temperature to fully dissolve

copper precursor. 0.5 g of tris(trimethylsily)silane (2 mmol) was added in to the solution

as reducing regent. The resulting clear blue mixture was heated up to 120 °C until the

solution turned into light yellow. Afterwards, the reaction temperature was slowly turned

up to and kept at 165 for 10 h while stirring. The color of the solution turned reddish

brown, indicating formation of copper nanowires. The product was harvested by

centrifugation at 6000 rmp for 5 mins. Then, the nanowires were washed repeatedly with

toluene using centrifugation-redispersion cycles to remove excess oleylamine. Finally,

the product was dispersed in toluene for further characterization and film fabrication.

Synthesis of silver nanowires1

Silver nanowires were synthesized following reported method in literature1. 0.34 g of

PVP (poly vinylpyrrolidone) was dissolved in 20 ml ethylene glycol in a reaction tube.

The solution was heated up and kept at 170 °C. Then, 0.025 g of AgCl was added to the

mixture upon vigorous stirring. After three minutes, 0.1 g of AgNO3 dissolved in

ethylene glycol was added drop wise to the mixture. The reaction was allowed to carry

out for another 30min when it was cooled down. Ag nanowires was centrifuged down

3

and washed with methanol 3 times. Finally, the nanowires were redispersed in methanol

for further characterization and film fabrication.

Fabrication of copper nanowire transparent conductor

To make a conductive thin film, copper nanowires were diluted using toluene and

sonicated for 15 min to form a homogenous suspension. The thin film was constructed by

filtering down the nanowires from the dispersion onto a nitrocellose porous membrane

(pore size 220 nm) via vacuum filtration. The nanowire network was transferred to a

transparent substrate (glass or PET) by applying pressure to the back side of the

membrane and forcing an intimate contact with the substrate. The copper nanowire thin

film was then annealed under forming gas (10% H2 and 90% Ar) at 200 for 30 min to

improve junction contact.

Fabrication of silver nanowire transparent conductor

Ag nanowires were diluted using methanol and sonicated for 15min to form a

homogenous suspension. The thin film was constructed by filtering down the nanowires

from the dispersion onto a polytetrafluorethylene (PTFE) membrane filter (purchased

from Sartorius, pore size 450nm). The nanowire network was transferred on to a

transparent substrate (glass) by applying pressure to the back side of the membrane.

Characterizations:

The morphologies of the as-grown copper nanowires were examined using a transmission

electron microscope (TEM, Hitachi H7650) and scanning electron microscope (SEM,

JOEL JSM - 6340F). The structure of copper nanowires was analyzed using high

resolution transmission electron microscopy (HRTEM, FEI Tecnai G20), selected area

electron diffraction (SAED) and X-ray diffraction pattern analysis (Bruker D8 Advance).

Exit-wave reconstruction was performed using MacTempas software. Sheet resistance of

nanowire thin film was measured using a four-point probe method (CDE-RESMAP-270).

The transmittance and haze measurement was carried out on an ultraviolet-visible near-

infrared spectrophotometer with an integrating sphere (Shimadzu UV-2550). AFM

images of the copper nanowires and their junction were taken using an Asylum MFP 3D

4

in tapping mode. The as-made Cu nanowire film on glass without heat treatment and the

one with 30 minutes forming gas annealing were used for the measurements.

Haze measurement

The haze measurement is carried out by D1003-13 standard2. Shimadzu UV-2550

ultraviolet-visible near-infrared spectrophotometer with an integrating sphere was used

for haze measurement. Four transmittance scans of a sample with different configurations

were acquired for its haze calculations: T1, incident light; T2, total light transmitted by the

specimen; T3, light scattered by the instrument and T4, light scattered by the instrument

and specimen. The haze factor of one specimen can be calculated by the equation:

Haze, % = [(T4/T2) - (T3/T1)].

Unless specified, all haze factors in discussion is the value measured at 550 nm of

wavelength. The contribution of glass substrate is already excluded.

References:

1. Hu, L., Kim, H. S., Lee, J., Peumans, P.; Cui, Y. ACS Nano 2010, 4, 2955–2963.

2. ASTM D1003-13, 2013. Standard Test Method for Haze and Luminous

Transmittance of Transparent Plastics, The American Society for Testing and

Materials, West Conshohocken, PA.

Figur

nanow

amou

dispe

in b)

netwo

Figur

after

re S1. The e

wires synthe

unt of oleic a

ersion. Copp

show severe

ork.

re S2. Nano

exposed in a

effect of olei

esize (a) with

acid in the re

er nanowire

e aggregation

wire surface

air for (a), 2

c acid as a s

h oleic acid (

eaction syste

synthesized

n which is n

e oxides. HR

hours and (b

5

econdary lig

(0.1 g), and

em can greatl

d with oleyla

not suitable f

RTEM image

b), 5 days. A

gand. TEM i

(b), without

ly improve c

amine as the

for the forma

es of the surf

A thin layer o

images of co

t oleic acid. S

copper nano

sole ligands

ation of loos

face of the C

of native oxi

opper

Small

wires’

s (as shown

e nanowire

Cu nanowire

ides is

obser

thickn

furthe

the in

Figur

coppe

cubic

rved on the a

ness is aroun

er growth of

nitial native o

re S3. Single

er product b

c copper nan

as grown cop

nd 2 to 3 nan

f oxide layer

oxide layer w

e crystalline

efore the app

noparticle

pper nanowi

nometer and

r was seen af

would stop f

e cubic coppe

pearance of

6

ires after was

d is consist of

fter 5 days ex

further oxida

er particle. (

five twinned

shing proced

f a mixture o

xposure in a

ation of the c

a) TEM ima

d particles. (

dure. The ox

of Cu2O and

ambient air. T

copper nano

age of the ma

(b) HRTEM

xide layer

d CuO. No

This implies

wire.

ajority of

M image of th

s

he

Figur

coppe

to (f)

With

metal

effect

conce

tende

oleyl

In the

are m

depos

later

the r

chara

chara

detail

re S4. The

er products w

), 0 g, 0.5 g,

presence o

l particles i

tive reducin

entration, th

ency of aniso

amine.

e absence of

mainly attribu

sition on pr

explanation

reaction prod

acterized by

acteristic inte

led high-res

role of oley

with differen

, 1 g, 2 g, 3

of non-coord

is also poss

ng regent in

he shape of

otropic grow

f “catalyzed

uted to two

referentially

is more pla

ducts were

TEM. We

ermediate of

solution TEM

ylamine in c

nt amount of

g, and 5 g

dinate solven

sible which

n this system

f copper pro

wth can also

growth”, on

possible rou

protected f

ausible for th

extracted fr

did not obse

f an oriented

M study on

7

copper nano

f oleylamine

of oleylamin

nt only, the

proves tha

m. With th

oducts beco

be seen corr

ne dimension

utes: oriente

facets. Acco

he following

from the rea

erve any ch

d attachment

n the as gro

owire shape

e within the r

ne were in p

e reduction

at tristrimeth

he increasing

omes better

related with

nal nanomat

d attachmen

ording to ou

g reasons. F

action and t

hain-like inte

t directed syn

wn nanowir

control. TE

reaction syst

presence of

of copper i

hylsiylsilane

g amount o

defined. A

the increasi

terial growth

nt and select

ur experimen

Firstly, differ

their morph

ermediates w

nthesis. Seco

re and it sh

EM images o

tem. From (a

the reactant

on to coppe

e acts as th

of oleylamin

An increasin

ing amount o

h mechanism

tive monome

ntal data, th

rent stages o

hologies wer

which are th

ondly, we di

hows that th

of

a)

ts.

er

he

ne

ng

of

ms

er

he

of

re

he

id

he

nanow

five (

follow

tips w

Third

we o

nanow

distri

by (1

Figur

nanop

coppe

weigh

produ

wire adopts

(111) facets

w the oriente

with one an

dly, there are

observed dur

wire growth

ibution, their

11) facets at

re S5. TEM

particles wit

er nanomate

ht. Nanowire

ucts.

a five-fold t

and it looks

ed attachmen

nother to for

e compelling

ring the ear

h. Besides t

r five-fold n

t the tip) are

image of na

th the dimen

erials (all mo

es take up ar

twinned stru

s like a five-

nt mechanism

rm a continu

g evidences

rly stage of

the fact that

ature and ex

all strongly

anoparticle b

sion ranging

orphology in

round 90% (

8

ucture. The t

fold pyramid

m, two segm

uous wire w

suggesting

f the reactio

t they have

xposed facets

correlated.

byproducts.

g from 30-10

cluded) of a

(by count) am

tips of the n

d. If the gro

ments would

which is not

that the five

on should se

strong corr

s (the nanop

The majorit

00 nm in dia

a typical synt

mong all mo

anowires are

owing mecha

d need to atta

t energetical

e-fold twin

erve as the

relation in

particles are

ty of byprodu

ameter. The y

thesis is arou

orphologies o

e bounded b

anism were t

ach pyramida

lly favorable

nanoparticle

seeds of th

tense of siz

also bounde

ucts are

yield of

und 50% by

of all copper

by

to

al

e.

es

he

ze

ed

r

Figur

imag

(b), 0

above

more

re S6. Tunin

es (Scale bar

0.5 g; and (c)

e three react

nuclei it wi

ng of nanowi

100nm) of c

), 0.25 g. (d)

tion conditio

ll generate, t

ire diameter

copper nanow

), diameter d

ons. During t

the smaller t

9

by changing

wire when th

distribution s

the nucleatio

the seeds.

g reducing a

he TTMSS a

statistics of c

on, the more

agent quantit

adding amou

copper nanow

reducing reg

ty. TEM

unt is (a), 1g;

wire in the

gent, the

;

10

Table S1. The effect of TTMSS quantity, copper precursor concentration and reaction

temperature on copper nanowire synthesis. The red highlighted condition is the optimized

reaction condition used for copper film fabrication.

CuCl2

(mmol)

TTMSS

(mmol)

Temperature

( ) Results

0.5 0 165 No Copper product

0.5 0.05 165 Copper cubes only

0.5 1 165 Nanowires with D ~ 22.5 ± 3.9 nm




1 4 165 Nanowires with D ~ 18.5 ± 4.7 nm

0.5 2 135 No copper product

When no TTMSS in the reaction system, copper chloride cannot be reduced to copper

metal. When small amount of TTMSS (10 % of CuCl2) is added, copper cubic

nanoparticles are observed in the product. However, the yield of Cu (0) is very low.

These imply that TTMSS does act as the effective reducing reagent of copper precursor,

and insufficient loading of silane cannot fully reduce CuCl2. The nanowire diameter can

be tuned by controlling the molar ratio of TTMSS/CuCl2. With set amount of CuCl2, the

nanowire diameter decreases as TTMSS increases. The size of the nanowire is also

sensitive about reactant concentration. Smaller concentration results in thicker nanowires.

The red highlighted condition is chosen as the optimized condition for typical nanowire

growth in this work for the reason that it gives less particle byproducts.

Figur

annea

16.98

summ

nanow

15.00

which

partia

re S7. AFM

aling. Before

8 nm. The m

mation of tw

wire film, th

0 nm which a

h is 5 nm sh

ally wield na

images of c

e annealing,

measured junc

o individual

he measured

add up to 29

ort of the su

anowire toge

copper nanow

the two mea

ction height

nanowire di

diameters o

9.50nm. The

ummation. Th

ether which c

11

wire film (a)

asure nanow

is 33.74 nm

iameters (33

f the two ind

annealed ju

his shows th

can enhance

), before ann

wires have di

m which is alm

3.64 nm). In

dividual wire

unction has th

hat the therm

e the junction

nealing and (

ameter of 16

most the sam

the case ann

es are 14.50

he height of

mal annealing

n contact.

(b), after

6.66 nm and

me with the

nealed

nm and

f 24.46nm

g can

d

Figur

is the

melt

resist

nanow

nanow

nanow

condu

Figur

most

and (

re S8. Depen

e optimized a

the nanowir

tance is obse

wire film is

wire. The tem

wire and dam

uctivity

re S9. SEM

nanowires a

c) 260 , w

ndency betw

annealing tem

e to form int

erved. When

almost insol

mperature ca

mage the nan

images of n

are intact; (b

when most of

ween film she

mperature. I

timate conta

n the annealin

lated due to t

annot be too

nowire perco

nanowire con

b) 230 , wh

f the nanowir

12

eet resistanc

If the temper

act at the wir

ng temperatu

the presence

o high either.

olation, whic

nducting film

hen some of

res are defor

ce and annea

rature is low

re-wire junct

ure is lower

e of oxides o

. Overheatin

ch will resul

m when anne

the nanowir

rmed.

aling tempera

wer, it cannot

tion. An incr

than 150

on the surfac

ng will sever

lt in degradin

ealed at (a) 2

res are visibl

ature. 200

t effectively

rease in shee

, the

ce of the

rely melt

ng in film

200 when

ly melted;

et

Figur

mech

polye

(MEL

of 80

angle

is app

when

extra

cycle

re S10. Con

hanical bendi

ethylene tere

LINEX® ST

0~90 ohms/sq

e is around 9

plied with m

n the sheet w

ordinary flex

es of bending

ductivity per

ing. Inset, fl

ephthalate (P

T506) were p

q) on PET (1

90 degree. IT

mechanical be

was bended 5

xibility. The

g.

rformance o

exible coppe

PET). Transp

purchased fro

175 micron)

TO on PET s

ending. The

00 cycles. In

e sheet resista

13

of ITO and C

er nanowire

parent PET (

om TEKRA

was purcha

hows severe

sheet resista

n compariso

ance shows

Cu nanowire

conducting

(175 micron)

A. ITO (sheet

ased from Sig

e degradation

ance increase

on, copper na

almost no ch

conductor u

film fabrica

) substrates

t resistance a

gma-Aldrich

n in conduct

ed almost 10

anowire film

hange even a

upon

ated on

at the range

h. Bending

tivity when i

00 times

m exhibit

after 1000

it

Figur

transp

Trans

suppo

The b

bendi

of ou

condu

were

proce

re S11. Shee

parent films

sparent PET

orting substr

bending angl

ing angle is

ur as made fi

uctivity rem

made from

ess. The four

et resistance

after applied

(175 micron

rate. Four be

le is 180 deg

90 degree of

lm. The film

ains its origi

the same bat

r films tests h

(normalized

d with multi

n) substrates

ending radii

gree for tests

f the test wit

ms show grea

inal value du

tch of coppe

have the she

14

d for clear co

iple bending

s (MELINEX

were tested

s with bendin

th bending ra

at resistance

uring the wh

er nanowires

eet resistance

omparison) p

g cycles with

X® ST506) w

: 8.2mm, 4.3

ng radius sm

adius of 8.2

towards me

hole experim

s and underw

e of 33 ± 3

performance

h various ben

were used a

3mm, 3.2mm

maller than 5

mm due to t

echanical ben

ment. All the

went the sam

ohms/sq.

e of flexible

nding radius.

s the

m and 2.0mm

mm. The

the small siz

nding and th

films tested

me fabrication

.

m.

ze

he

n

Figur

sized

Vacu

Figur

repre

keep

the co

re S12. SEM

d nitrocellulo

uum filtration

re S13. Stab

sents a diffe

the high con

onductivity o

M images of

ose membran

n can effecti

bility test of c

erent loading

nductivity pe

of the coppe

(a) copper n

nes and (b) c

vely remove

copper nano

g amount. It i

ermanently.

er film degra

15

nanowires wh

copper nanow

e most of the

owire transpa

is also worth

After severa

ades significa

hich were fil

wires drop c

e nanoparticl

arent conduc

h to note tha

al months’ ex

antly.

ltered by 22

asted on sili

le byproduct

ctors. Each c

at the Cu NW

xposure in a

0 nm pore-

con wafer.

ts.

color

W film canno

ambient air,

ot

Figur

Figur

transp

total

re S14. Cop

re S15. Visu

parent condu

transmittanc

per film is fu

ual comparis

uctor and an

ce.

fully oxidized

son between

ultra-thin co

90

16

d after 3 mo

a silver nan

opper nanow

days in air

nths storage

nowire (with

wire conduct

e in air

mean diame

ting film wit

eter of 50nm

th the same

m)

Synthesis of ultra-thin copper nanowires using tris...

Documents

Transcript of Synthesis of ultra-thin copper nanowires using tris...