semiconductor nanowires directly grown on graphene – towards wafer scale transferable nanowire...
TRANSCRIPT
Nanoscale
COMMUNICATION
Dow
nloa
ded
by M
cGill
Uni
vers
ity o
n 07
/04/
2013
12:
08:5
9.
Publ
ishe
d on
21
Mar
ch 2
013
on h
ttp://
pubs
.rsc
.org
| do
i:10.
1039
/C3N
R00
367A
View Article OnlineView Journal
aDepartment of Chemical and Biomolecular
Berkeley, Berkeley, CA, 94720 USA. E-mail:bCenter of Integrated Nanomechanical Syste
Berkeley, CA, 94720, USA
Cite this: DOI: 10.1039/c3nr00367a
Received 21st January 2013Accepted 20th March 2013
DOI: 10.1039/c3nr00367a
www.rsc.org/nanoscale
This journal is ª The Royal Society of
Semiconductor nanowires directly grown on graphene –towards wafer scale transferable nanowire arrays withimproved electrical contact
John P. Alper,ab Albert Gutes,ab Carlo Carraroa and Roya Maboudian*ab
We present for the first time the growth of dense arrays of silicon
and silicon carbide nanowires directly on graphene as well as
methods of transferring these novel hybrids to arbitrary
substrates. Improved electrical contact for SiC nanowire/graphene
hybrid is demonstrated in the application of a robust supercapacitor
electrode.
1 Introduction
The eld of nano-electronics is rapidly evolving as new anduseful scale-dependent properties associated with nano-mate-rials are discovered. Arrays of semiconducting nanowires areparticularly attractive for a number of applications. Insecondary battery technology, silicon nanowires (SiNWs) havebeen shown to reduce the mechanical stresses associated withlithium insertion and extraction.1 This may enable rechargeablebatteries with over ten times the specic capacity of currentgraphite anodes. SiNW arrays also exhibit greatly enhancedbroadband light absorption due to decreased reectance andtransmittance as compared to solid lms,2 and are promising inthe eld of thin lm photovoltaics. In addition to Si, thenanoscale electrical and morphological properties of siliconcarbide NWs (SiCNWs) including high aspect ratio and lowelectron affinity indicate them as excellent candidates for eldemission cathodes.3 To date, SiCNW emitters have beendemonstrated with a lower turn-on eld and higher currentdensities than Si based devices.4 SiCNWs also show promise asstable materials in electrochemical environments as encoun-tered in aqueous supercapacitors.5
Typically these semiconducting nanowire arrays are grownfromrigid conductive substrateswhichprovide electrical contactto the array. A exible conductive substrate however providesmany advantages in terms of the applications described above.
Engineering, University of California at
ms, University of California at Berkeley,
Chemistry 2013
Batteries and supercapacitors both benet from smaller formfactors such as the typical rolled cylinder cell. The labor andmounting materials for solar panel installation, �15% of thetotal cost per Watt,6 may be signicantly reduced by “roll out”light-weight modules. Flexibility also provides greater opportu-nities for integration of energy storage and harvesting technol-ogies with fabric for use in clothing or lightweight buildingsupplies. In order to realize these benets, methods must bedeveloped for transferring arrays of nanowires to exiblesubstrates while maintaining good electrical contact.
Considering the latter issue rst, forming electrical contactto nanomaterials is a non-trivial task which has received muchattention. Optical lithography,7 electron beam lithography,8 dip-pen nanolithography,9 and focused ion beam methods,10 havebeen used to contact individual nanowires, nanorods, andnanotubes; however, these are not applicable to arrays ofnanomaterials at scale. Evaporation of conductors onto the tipsof nanowires is a method which addresses a larger array ofnanomaterials, but requires a polymer deposition to preventshorting, and partial removal of the polymer to expose thenanowires.11
The post-growth transfer of nanomaterials arrays has been aswell the subject of much study. Polymers have been utilized toform a supportive matrix around vertically aligned nanowirearrays which are then transferred to a secondary substrate aermechanical removal from the initial growth substrate. Theresulting nanowires may be in a vertical,12,13 or horizontalorientation.14 Electrical contact is then made to the array bycontact evaporation. However, the application of polymer mayrequire a later polymer removal step unless the polymer isincorporated into the device architecture. Wet approaches suchas the Langmuir–Blodgett technique are an alternative methodof depositing arrays of nanomaterials on substrates.15,16 Thisobviates the need to stabilize the entire array prior to transfer,as the materials need only to be dispersed on the surface of adeposition liquid. Still the resulting material lacks precisioncontrol of electrical contact points and thus it is not favourablefor scalable, array-based devices.
Nanoscale
Nanoscale Communication
Dow
nloa
ded
by M
cGill
Uni
vers
ity o
n 07
/04/
2013
12:
08:5
9.
Publ
ishe
d on
21
Mar
ch 2
013
on h
ttp://
pubs
.rsc
.org
| do
i:10.
1039
/C3N
R00
367A
View Article Online
Here we demonstrate a simplied route for the transfer ofelectrically contacted semiconductor nanowire arrays, whereindense arrays of semiconducting nanowires are grown directlyon graphene which may then be transferred to an arbitrarysubstrate. We have chosen graphene due to its exibility,excellent electrical and optical properties,17,18 as well as itsthermal stability at the utilized nanowire growth tempera-tures.19 Transfer techniques for graphene have also beenpreviously developed,20,21 which provided a process frameworkwhich we have further optimized in this work.
Fig. 1 SiCNW on graphene transferred onto a flexible plastic substrate.
2 Experimental2.1 Graphene growth
Graphene is grown by chemical vapor deposition on copper asdescribed previously.19 In short, an 8 � 2 cm2 99.998% copperfoil of 0.025 mm thickness (Alfa Aesar, USA) is introduced into aquartz tube and heated under vacuum to 1000 �C at a rate of�55 �C min�1 with a 10 sccm of H2 ow (Praxair 99.999%). Cufoil is annealed under these conditions for 30 minutes. Then, 20sccm of methane (Praxair 99.993%) is introduced and main-tained for 15 minutes. The samples are cooled at a rate of�15 �C min�1 under the same ambient. When the temperaturereaches 600 �C all gas ows are stopped and the samples arecooled to room temperature under vacuum (�20 mTorr). Thegraphene obtained by this method is typically monolayer withsmall areas of two or three layer patches.
2.2 Nanowire catalyst deposition
Graphene decorationwith Au nanoparticles for the SiNWgrowthis performed on the copper/graphene foil via electroless depo-sition as described in a previous work.22 Graphene on Cu foil isimmersed in a 0.2mMKAuCl4 (Sigma-Aldrich, 98%) solution for30 seconds, rinsed in de-ionized water and dried under N2.
Nickel deposition on graphene for the SiCNW growth occursaer transfer of the graphene onto the sacricial SiO2 support.The thin lm,�2.2 nm of Ni, is deposited onto the graphene viae-beam evaporation using a Thermionics VE-700 VacuumEvaporator. The bare SiO2 region surrounding the graphene ismasked with Teon tape to ensure it remains Ni free.
2.3 Graphene transfer to NW growth substrate
Graphene grown on Cu foil is transferred onto a 100 nm SiO2
lm on Si wafer using a method described previously.20 Brieypolymethylmethacrylate (PMMA), 8% wt/wt in toluene, is spincoated onto the copper/graphene foil which is subsequentlyoated on a 1 M FeCl3 (Cu etching) solution. Aer etching ofthe Cu, the graphene/PMMA composite is gently lied out ofthe solution with a glass slide, rinsed with DI water andallowed to dry in air on the SiO2 substrate. The PMMA is thenremoved by dissolution in toluene, an improvement from thereferenced procedure. It was found that the slower etching rateof PMMA in toluene, as compared to acetone, reduces theincidence of the lm curling onto itself, which we attribute touneven stresses being released in a rapid manner duringacetone dissolution.
Nanoscale
2.4 Nanowire growth
SiNWs are synthesized via a vapor–liquid–solid mechanism asdiscussedelsewhere.23Graphene/AuNPonSiO2 areplaced into anatmospheric pressure chemical vapor deposition (CVD) furnaceand heated to 850 �C, at a rate of �57 �C min�1, under a ow of200 sccm 10%H2 in Ar. Then 50 sccm of SiCl4 (Strem Chemicals,99.9999%) is added to thedirectowandheld for 5minutes. Aergrowth, theSiCl4ow is stoppedand the reactor is cooleddown to�200 �C before removing the graphene/SiNW sample.
SiC nanowires are grown from the graphene substrate asdescribed in earlier work.24 SiCNW's are grown from the Nicatalyst in a low-pressure CVD furnace. Samples are heated to950 �C, at a rate of �55 �C min�1 under 10 sccm of H2. Uponreaching the growth temperature, the ow of hydrogen isreduced to 5 sccm and a 0.5 sccm ow of methyltrichlorosilane(MTS, Sigma-Aldrich, 99%) is introduced to the chamber.Growth is allowed to proceed for 30min, at which point theMTSow is stopped and the sample is cooled to room temperature.These conditions have been shown to result in a nanowire arrayof approximately 5 mm height.5
2.5 Nanowire/graphene hybrid transfer to arbitrarysubstrate
The nanowire arrays are transferred onto an arbitrary substrateby a simple sacricial etch transfer. The SiO2 lm is rstetched in hydrouoric acid (48%) for �10 minutes. Thesamples are carefully removed from the solution with the NW/graphene orientated upwards to prevent slipping of the gra-phene from the Si substrate. The NW/graphene/Si substratesare then slowly introduced to a DI bath at a shallow angle. Theinteraction between the hydrophobic graphene and watersurface results in the NW/graphene lm oating on thesurface. It is then removed from the surface of the water bathby an arbitrary substrate. In this study, sapphire is used as thesubstrate to place the SiNW/graphene samples for XRD anal-ysis and a thin plastic lm (3M AF4300 transparency lm) isused to deposit SiCNW/graphene samples for optical imagingin Fig. 1.
This journal is ª The Royal Society of Chemistry 2013
Communication Nanoscale
Dow
nloa
ded
by M
cGill
Uni
vers
ity o
n 07
/04/
2013
12:
08:5
9.
Publ
ishe
d on
21
Mar
ch 2
013
on h
ttp://
pubs
.rsc
.org
| do
i:10.
1039
/C3N
R00
367A
View Article Online
2.6 Characterization methods
Nanowire morphology is characterized by scanning electronmicroscopy (SEM) using a NovelX mySEM. Crystal structure isdetermined using a Siemens D5000 automated X-ray diffrac-tometer (XRD) operated in q–2q geometry. Graphene qualitydetermination is performed by Raman spectroscopy (HoribaJYLabRAM). Electrochemical testing of the samples is performedusing cyclic voltammetry (CHI 660D electrochemical station) in1 M KCl utilizing a three-electrode cell consisting of an Ag/AgClreference electrode, Pt wire counter electrode and the NW array/graphene hybrid as the working electrode. Electrical connectionis made directly to the front face of the working electrode by Agepoxy paste deposited into a small region of the SiCNW array.Capacitance, C, is calculated using C ¼ (i/A)(dt/dV), where i isthe current at V ¼ 0.2 V during the positive sweep, dV/dt is thescan rate used for the measurement, and A is the projectedsurface area of the sample immersed in the solution.
Fig. 2 SEM of the obtained nanowires on graphene: silicon NWs (a) and (b).Scale bars are 10 and 2 mm respectively. Silicon carbide NWs (c) and (d). Scale barsare 10 and 2 mm respectively.
Fig. 3 (a) Raman spectrum of graphene after SiNW growth, using laser excita-tion wavelength 633 nm. (b) XRD spectrum of the SiNW grown on graphene withfinal transfer on a sapphire substrate.
3 Results and discussion
The generality of the nanowire growth procedures used in thiswork indicates that the overall method of growth and transfer,described above and summarized in Scheme 1, may be extendedto a wider array of semiconducting nanomaterials grown withsimilar methods. While SiO2 is chosen here as the sacriciallayer, the method can be extended to the use of any number ofother lms so long as they withstand the growth conditions forthe nanowires and may be etched in a facile manner while notattacking the desired synthesized materials. The hydrophobicnature of graphene is exploited in oating the NW lms on topof the DI rinse bath aer HF etching. This method works mostconsistently with the SiCNW arrays, which are themselveshydrophilic as synthesized (water contact angle of �0�). TheSiNWs however, being quite hydrophobic aer exposure to theHF etching solution, have a tendency to self-associate. Thisresults in the SiNW/G hybrid curling up on itself when placedinto the DI rinse bath. It is found that these hybrids are mostconsistently transferred intact if the SiNW array is not allowedto dry in between removal from the HF bath and placement inthe DI rinse bath.
Scanning electron microscopy (SEM) analysis of both typesof NW/G hybrids aer transfer are presented in Fig. 2. A verydense lm of nanowires is achieved for both cases in contrast toprevious work on InAs semiconductor nanowire growth fromgraphene.25 This is a result of the catalyst seeding and NW
Scheme 1 Graphene growth, decoration and transfer process followed by Si and SiFor SiNW growth, (b.1) graphene is then decorated with gold nanoparticles by galvagraphene is transferred to a SiO2 substrate and then (c.1) coated with an evaporatenanowire/graphene film is then transferred to an (e) arbitrary substrate.
This journal is ª The Royal Society of Chemistry 2013
growth technique used here whereas the previous work utilizeda catalyst free method. Because the catalysts are on top of gra-phene on amorphous oxide, no preferential growth directionfrom the substrate is observed for either type of nanowire. ARaman spectrum taken from the same sample in an Au-freeregion, shown in Fig. 3a, presents a negligible D peak and isconsistent with the published Raman spectra for high quality
C nanowire growth and final transfer. First, (a) graphene is grown on a copper foil.nic displacement and (c.1) transferred to a SiO2 substrate. For SiCNW growth, (b.2)d Ni film. (d) Nanowires are then grown on the graphene via metal catalysed. The
Nanoscale
Fig. 4 (a) Comparison between cyclic voltammetry results for SiCNWs grown on graphene vs. those grown on oxide at a scan rate of 500 mV s�1. (b) Cyclic vol-tammetry results for SiCNW/G hybrid material over a range of scan rates. Rates are in V s�1. (c) Retained capacitance of the SiCNW vs. number of cycles.
Nanoscale Communication
Dow
nloa
ded
by M
cGill
Uni
vers
ity o
n 07
/04/
2013
12:
08:5
9.
Publ
ishe
d on
21
Mar
ch 2
013
on h
ttp://
pubs
.rsc
.org
| do
i:10.
1039
/C3N
R00
367A
View Article Online
graphene grown with the copper CVD method utilized here.19
This indicates that the graphene is stable during the catalystdecoration process, transfer and NW growth. The siliconnanowires are highly crystalline as observed from the sharpX-ray diffraction (XRD) spectrum in Fig. 3b taken aer transferof the SiNW/G hybrid to a sapphire substrate.
As an example of the utility of the NW/G hybrid material, andto demonstrate that good electrical contact between the NWarray and the underlying graphene support is achieved, thehybrid material is investigated as an electrode for aqueoussupercapacitor applications. A SiC nanowire/graphene sheet onSiO2 is probed using cyclic voltammetry (CV) over a wide rangeof scan rates. Results for a SiCNW array grown directly on a thinlm of SiC on SiO2 using the same technique and growth timeare used as a comparison. Electrical connection is achieved bythe application of epoxy-Ag paste on the arrays. Cyclic voltam-metry results indicate that the SiCNW on graphene arrays havesimilar capacitance values, �350 mF cm�2, as SiCNWs grown ona thin lm of SiC, �370 mF cm�2. This indicates that the entirearray is electrically connected. Fig. 4a shows the cyclic voltam-mograms of the SiCNW arrays grown on graphene compared tothose grown on SiO2. A very different behaviour at a scan rate of0.5 V s�1 can be observed. While the SiCNW/graphene hybriddisplays a near ideal capacitive rectangular-shaped current–voltage relationship, the SiCNW/SiO2 results are quite skewed,indicative of higher internal resistances. These resistances arisefrom the conductivity of the wires, the contact between thewires and the substrate, the resistance through the substrateand the contact resistance where electrical connection is madebetween the testing station and the NW electrode. As thenanowires are grown from the same precursor and catalystunder identical conditions, and connection to the array is madevia the same silver epoxy, it is concluded that the resistance atthe wire–substrate interface and the resistance throughsubstrate are the main contributing factors to the differencebetween the CV behaviours of the two tested materials. Thecontact between the nanowires grown in the absence of gra-phene may be due to the physical contact between the wires inthe nest-like morphology. In contrast, the NW/graphenecomposite, with greatly enhanced conductivity through thearray, is expected to be in contact through the graphene underlayer.
The extent of this difference can be qualitatively demon-strated with the same technique by considering the results for
Nanoscale
cyclic voltammetry on the SiCNW/graphene over a wider rangeof scan rates, as shown Fig. 4b. The SiCNW/graphenehybrid maintains near ideal performance through a scan rate of2 V s�1. It has been shown that graphitic interlayers can reducethe contact resistance between the deposited SiC thin lms andmetal contacts,26 and thus it seems reasonable that withimproved graphene growth and transfer techniques, a grapheneinterlayer may be used to the advantage of the electronicmaterials engineer when designing hybrid conductive struc-tures. The robustness of the NW/G hybrid material is conrmedby performing 10 000 complete charge–discharge cycles in anaqueous electrolyte at a scan rate of 5 V s�1. Fig. 4c shows thenegligible reduction in capacitance measured aer this time,indicating high stability during electrochemical cycling.
Electrical conductivity measurements on SiCNW/graphenehybrid material post transfer to a exible substrate have beenperformed using a two-probe conguration. The results indicatethat the conductivity is not greatly affected by the transfer step,remaining on the order of 10 kU for a sample size of �1 cm � 1cm. More rigorous conductivity measurements under differentexing regimes are currently underway.
4 Conclusions
In summary silicon and silicon carbide nanowires have beensuccessfully grown directly on graphene and transferred ontoarbitrary substrates. Silicon nanowires are grown by metal cat-alysed CVD using gold deposited on graphene by an electrolessdeposition before PMMA transfer. SiC nanowires are synthe-sized with a similar process using evaporated nickel on top ofgraphene aer PMMA transfer. SEM conrmed the nanowirestructure of both semiconductors on graphene. Raman spec-troscopy revealed graphene stability aer NW growth whileelectrochemical measurements revealed the excellent contactbetween SiC nanowires and graphene, opening the possibility ofusing this new hybrid material for energy storage devicesamong others. Transfer of arrays of these semiconducting wiresto arbitrary substrates is demonstrated.
Notes and references
1 C. K. Chan, H. Peng, G. Liu, K. McIlwrath, X. Fengzhang,R. A. Huggins and Y. Cui, Nat. Nanotechnol., 2008, 3, 31.
This journal is ª The Royal Society of Chemistry 2013
Communication Nanoscale
Dow
nloa
ded
by M
cGill
Uni
vers
ity o
n 07
/04/
2013
12:
08:5
9.
Publ
ishe
d on
21
Mar
ch 2
013
on h
ttp://
pubs
.rsc
.org
| do
i:10.
1039
/C3N
R00
367A
View Article Online
2 L. Tsakalakos, J. Balch, J. Fronheiser and B. A. Korevaar,Appl. Phys. Lett., 2007, 91, 233117.
3 K. Zekentes and K. Rogdakis, J. Phys. D: Appl. Phys., 2011, 44,133001.
4 E. Spanakis, J. Dialektos, E. Stratakis, V. Zorba, P. Tzanetakisand C. Fotakis, Phys. Status Solidi C, 2008, 5, 3309.
5 J. P. Alper, M. S. Kim, M. Vincent, B. Hsia, V. Radmilovic,C. Carraro and R. Maboudian, J. Power Sources, 2013, 230,298.
6 R. Wiser, G. Barbose and C. Peterman, Tracking the Sun: TheInstalled Cost of Photovoltaics in the U.S. From 1998–2007,Lawrence Berkeley National Lab Internal Report, 2009,LBNL-1516e.
7 M. E. T. Molares, E. M. Hohberger, C. Schaeein, R. H. Blick,R. Neumann and C. Trautmann, Appl. Phys. Lett., 2003, 82,2139.
8 M. Bockrath, D. H. Cobden, P. L. McEuen, N. G. Chopra,A. Zettl, A. Thess and R. E. Smalley, Science, 1997, 275, 1922.
9 W. M. Wang, M. C. LeMieux, S. Selvarasah, M. R. Dokmeciand Z. Bao, ACS Nano, 2009, 3, 3543.
10 T. W. Ebbesen, H. J. Lezec, H. Hiura, J. W. Bennet,H. F. Ghaemi and T. Thio, Nature, 1996, 382, 54.
11 A. R. Abramson, W. C. Kim, S. T. Huxtable, H. Yan, Y. Wu,A. Majumdar, C. L. Tien and P. Yang, J. Microelectromech.Syst., 2004, 13, 505.
12 M. C. McAlpine, H. Ahmad, D. Wang and J. R. Heath, Nat.Mater., 2007, 6, 379.
13 J. M. Spurgeon, S. W. Boettcher, M. D. Kelzenberg,B. S. Brunschwig, H. A. Atwater and N. S. Lewis, Adv.Mater., 2010, 22, 3277.
This journal is ª The Royal Society of Chemistry 2013
14 Y.-K. Chang and F. C.-N. Hong, Nanotechnology, 2009, 20,195302.
15 D. Whang, S. Jin, Y. Wu and C. M. Lieber, Nano Lett., 2003, 3,1255.
16 L. Mai, Y. Gu, C. Han, B. Hu, W. Chen, P. Zhang, L. Xu,W. Guo and Y. Dai, Nano Lett., 2009, 9, 826.
17 K. S. Novoselov, A. K. Geim, S. V. Morozov, D. Jiang,Y. Zhang, S. V. Dubonox, I. V. Grigorieva and A. A. Firsov,Science, 2004, 306, 666.
18 R. R. Nair, P. Blake, A. N. Grigorenko, K. S. Novoselov,T. J. Booth, T. Stauber, N. M. R. Peres and A. K. Geim,Science, 2008, 320, 1308.
19 X. Li, W. Cai, J. An, S. Kim, J. Nah, D. Yang, R. Piner,A. Velamakanni, I. Jung, E. Tutuc, S. K. Banerjee,L. Colombo and R. S. Ruoff, Science, 2009, 324, 1312.
20 A. Reina, H. B. Son, L. Y. Jiao, B. Fan, M. S. Dresselhaus,Z. F. Liu and J. Kong, J. Phys. Chem. C, 2008, 112, 17741.
21 K. S. Kim, Y. Zhao, H. Jang, S. Y. Lee, J. M. Kim, K. S. Kim,J.-H. Ahn, P. Kim, J.-Y. Choi and B. H. Hong, Nature, 2009,457, 706.
22 A. Gutes, B. Hsia, A. Sussman, W. Mickelson, A. Zettl,C. Carraro and R. Maboudian, Nanoscale, 2012, 4, 438.
23 G. Doerk, N. Ferralis, C. Carraro and R. Maboudian, J. Mater.Chem., 2008, 18, 53761.
24 H.-J. Choi, H.-K. Seong, J. C. Lee and Y. M. Sung, J. Cryst.Growth, 2004, 269, 472.
25 Y. J. Hong, W. H. Lee, Y. P. Wu, R. S. Ruoff and T. Fukui,Nano Lett., 2012, 12, 1431.
26 F. Liu, B. Hsia, C. Carraro, A. P. Pisano and R. Maboudian,Appl. Phys. Lett., 2010, 97, 262107.
Nanoscale