2012-nl-Highly Efficient Charge Separation and Collection across in Situ

Highly Efficient Charge Separation and Collection across in Situ Doped Axial VLS-Grown Si Nanowire p?n Junctions

A.D.Mohite,*,?,§D.E.Perea,?,?,∥S.Singh,?,§,⊥S.A.Dayeh,?,?I.H.Campbell,?S.T.Picraux,*,?,?and H.Htoon*,?,§

?Center for Integrated Nanotechnologies,?Material Physics and Applications Division,and§Chemistry Division,Los Alamos National Laboratory,Los Alamos,New Mexico87544,United States

*Supporting Information

diffusion lengths,with surface recombination velocities(S)

previously for as-grown nanowires,all of which used hot wall reactors.

nanowires is comparable to or better than that achieved for nanowires

the dramatically improved surface recombination velocities may

our cold wall CVD reactor.

n junction,nanowire,scanning photocurrent microscopy

ilicon(Si)is an excellent material for solar energy applications as it offers a near ideal band gap for conversion of solar photons to electricity and is also widely abundant. However,the high cost of producing high purity wafers for single crystalline cells and the lower efficiency for polycrystal-line and amorphous thin film cells has restricted the more widespread use of silicon-based photovoltaic technology1?3and has promoted research for alternative nonplanar based Si cells such as Si nanowire(NW)solar cells.4?12Silicon NWs grown by the vapor?liquid?solid(VLS)process13offer distinct advantages for use in solar cells,including a reduced amount of Si material,control over the diameter and aspect ratio14for enhanced optical absorption,9,11,15ability to fabricate axial16 and radial p?n junction devices,17,18and in situ incorporation of dopants.19However the overall efficiency of photovoltaic devices fabricated to date from Si NW arrays has been limited to～7.9%,which is still far from predicted maximum values and much less than the15?20%efficiency desired for commercially viability.20A better understanding and control over the competition between charge recombination and charge separation/collection processes in these NWs is essential for designing more efficient Si NW solar cells.

The recombination of photogenerated minority carriers occurs mainly via dopants and a variety of crystalline defects which limits the charge separation and collection efficiency in bulk Si.In addition,minority carrier recombination in VLS grown NWs is expected to be further increased by the unintentional incorporation of metallic impurities from the growth catalysts(e.g.,Au),21as well as the increased influence of the relatively large proportion of unpassivated surface states.15The efficiency of the separation and collection of photogenerated carriers can be measured by the average distance or time a photogenerated carrier can travel before recombination,i.e.,minority carrier diffusion length,L,or lifetime,τ,which are related to one another by L=(Dτ)1/2, where D is the carrier diffusion coefficient.22In addition,the surface recombination velocity,S,provides a measure of the degree of surface passivation without the influence of dopant concentration or the increased surface to volume ratio with decreasing NW diameter.Experiments to date have applied electron beam induced current(EBIC)23or scanning photo-current microscopy(SPCM)24?26to measure the effective minority carrier diffusion lengths(L eff)across Schottky junctions of doped nano or microscale wires or ex situ doped

Received:December21,2011

Revised:March6,2012

Published:March20,2012

p?n junctions of micropillars.These studies have revealed that L eff values scale with the NW diameter.More importantly,they also revealed a dramatic increase in L eff by surface passivation via several different approaches such as with hydrogen termination or silicon nitride,26thin amorphous Si layer coatings,24or via background illumination.27These findings provide clear evidence that the recombination of photo-generated carriers at unpassivated surface states plays a dominant role in limiting the efficiency of the charge separation and collection processes.The ability to control the quality of the surface with CVD growth conditions and processing parameters therefore holds the key to achieving high perform-ance NW-based solar photovoltaics.

In this letter,we describe the first report on measuring both the electron and hole effective minority carrier diffusion lengths in in situ doped,Au catalyzed small diameter(<100nm)axial p?n junction Si NWs.In contrast to the Si NWs grown by the VLS method by previous groups using hot wall reactors,these p?n junction NWs were grown using a cold wall reactor in a hydrogen rich environment.This could significantly alter the reactivity of the precursor gases,for example minimizing NW sidewall reactions and deposition,and result in surfaces with a lower density of unpassivated surface states compared to those grown in hot wall reactors.Indeed,we observed that L eff for both the electrons and holes in our as-grown NWs are an order of magnitude longer than those previously reported on similar diameter Si NWs,23and even longer than was achieved after passivation with a thin amorphous Si layer.24Furthermore,we also observed that background light illumination,which was previously shown to optically passivate surface states and give rise to a dramatic increase in L eff of hot wall reactor grown microwires,27has no effect on our NWs.On the basis of these observations,we suggest that the density of surface states in NWs grown using a cold wall CVD reactor can be significantly lowered compared to hot-wall reactor growth.

The cold wall CVD reactor growth of silicon p?n junction

NWs used for this study employed SiH4as the Si source precursor and H2as the carrier gas.The NWs were grown using the VLS growth technique from a2nm thick UHV electron beam evaporated Au film deposited on a solvent cleaned and dilute HF etched and passivated Si(111)substrate28that was initially solvent cleaned and native oxide etched.Nanowire growth was initiated at550°C by simultaneously flowing300 and10sccm of SiH4(50%in H2)and B2H6(100ppm in H2), respectively.After2min of growth of the B-doped Si segment, a P-doped segment was grown by simultaneously stopping the B2H4flow and introducing PH3(100ppm in H2)at100sccm for3.5min while maintaining a300sccm SiH4flow.A total pressure of3Torr was maintained throughout the growth.The growth times were chosen such that the resulting NW lengths were～15μm with the p?n junction at the center of the NW. Devices were fabricated from individual NWs by first drop casting NWs suspended in isopropyl alcohol onto a100nm thick oxide layer on a degenerately doped Si substrate.The NW locations were mapped using an optical microscope,and electron beam lithography was used to define the contact areas. E-beam evaporation was then used to deposit120nm thick Ni contacts on both the B-doped and P-doped Si segments.Figure 1a shows a schematic of the device structure and an optical image of the actual device used for our SPCM measurements. Figure1b shows the current?voltage characteristics of a two terminal p?n diode device in the dark measurement(red curve)and under illumination(green curve).Upon optical excitation of the p?n junction with a green diode pumped solid state laser(λ=532nm)coupled through a confocal microscope,a typical photovoltaic response is observed with V oc=0.42V and I sc=65nA(green curve).We estimate an internal quantum efficiency(IQE)of～50%at zero bias for illumination with monochromatic light ofλ=532nm,a NW diameter=80nm,laser spot size～500nm,incident power= 100μW,absorption coefficient=104cm?1,and a reflection coefficient=0.35.The magnitude of the IQE suggests that the NW p?n junction exhibits strong photosensitivity,indicating good crystalline quality and photovoltaic properties. Through analysis of the spatial variation in the locally excited photogenerated electron?hole pairs measured by SPCM we determine(1)the nature of the contacts(ohmic or Schottky), (2)individual segment electrical type(p or n),and(3)the minority carrier diffusion lengths in our in situ doped Si p?n junction NWs.A schematic of the experimental setup for the measurement of the L eff is illustrated in Figure2a.A modulated green laser was coupled through a confocal microscope to obtain a diffraction limited laser spot which was scanned over a distance of approximately6μm between the two electrodes of the two terminal device in the direction from the p-doped region to the n-doped region of the NW,while measuring the photocurrent in200nm steps.At each position the photocurrent was measured as a function of applied bias which was scanned from+1to?1V with0.1V

increments. Figure1.(a)Schematic of the axial p?n junction device and an optical image of the actual device with ohmic contacts.(b)Current voltage curve measured in the dark(red)and under illumination(green)with 532nm excitation wavelength shows typical solar cell response.

The photocurrent was collected using a lock-in amplifier synched at the modulation frequency of the incident laser (513Hz).Figure 2b shows the photocurrent versus the position along the NW axis for the case of 0V applied bias.A peak in the photocurrent is observed at a position corresponding to approximately the middle of the NW,consistent with the intended location of the p ?n junction determined from the growth rate and sequence of the exposed dopants gases flown.As prior studies have reported high photocurrent near Schottky contacts,23our observation of a very small photocurrent near the contacts for our devices provides a clear indication of net ohmic contact behavior.As an additional check we also examined the case where we formed metal ?NW Schottky contacts and observed higher photocurrents near the NW ?metal contact as compared to the magnitude of the photo-current at the p ?n junction near the center of the NW device.From the polarity of the current flow,the known growth sequence and the direction of taper in the NW,we can unambiguously identify each of the p and n segments.The approximate doping concentrations in the boron doped p-and phosphorus-doped n segments were determined by standard four probe resistivity measurements performed separately on each segment of the NW grown in the same batch as the one discussed here.Under the assumption of bulk mobility values we obtain N A ～2×1016cm ?3for holes and N D ～2×1018cm ?3for electrons,indicating a p ?n +doping sequence for our axial NW junctions.

The peak in the photocurrent signal corresponds to the geometric center of the depletion region (see Figure 2b),with the photocurrent expected to decay exponentially on either side.To understand this,consider the schematic band diagram shown in Figure 2c and corresponding space charge region in Figure 2d representing a graded p ?n +junction with the larger depletion width on the p side of the junction.The charge collection probability is maximum when the electron ?hole pairs are generated inside or within a diffusion length of the depletion region where they are separated by the built-in

electric field and drift toward their respective contacts to generate a photocurrent.Upon moving away from the depletion region,the collection probability and thus the photocurrent decays exponentially with distance,with different decay rates for electrons and holes.Beyond a distance of one diffusion length,the generated minority carriers have a higher probability to recombine with the majority carriers of opposite polarity than to be collected.

The measured L eff for both electrons and holes within the in situ doped p ?n junction nanowires described here are an order of magnitude greater than previously reported for similar diameters.Figure 3a shows the semi log plot of the photocurrent at 0V applied bias (open circles)for a p ?n junction NW of diameter 70nm.The vertical dashed lines mark the edge of the hole (blue)and electron (red)collector contacts.The effective minority carrier diffusion length for

electrons (L e eff )and holes (L h

eff

)was extracted by fitting the exponential decay of the photocurrent to the equation I =I 0

exp(±(x ?d )/L e,h eff

),as described previously by Gu et al.29Here

we find a good fit and determine a value of L e

eff

≈1.0μm for electrons and L h eff

≈0.66μm for holes.As an additional check

on these surprisingly long L e,h

eff

in Si NWs,we performed SPCM measurements on a n-type NW segment for which a Schottky barrier contact with Ni was formed.30As seen in Figure 3b,the photocurrent versus position shows a maximum photocurrent at the contact position (vertical dashed line)where the photogenerated electron ?hole pairs are separated by the depletion region formed between the NW and the Ni contact.

A fit to the exponential decay in this case gives L h

eff

≈0.64μm for the NW with d ≈70nm,in good agreement with the results obtained for the p ?n junction devices.In order to eliminate the possibility of the measured diffusion lengths being limited by the spatial resolution of the SPCM measurement,other control experiments included using a smaller excitation spot size,as well as making SPCM scans perpendicular to the NW length were performed and can be found in the Supporting

Information.

Figure 2.(a)Experimental setup for SPCM https://www.wendangku.net/doc/fc12852641.html,ser is scanned along the nanowire and photocurrent is collected as a function of applied dc bias.(b)Photocurrent profile along the length of the nanowire with blue half of the curve representing electron transport and red represents hole transport.(c)Schematic of the band diagram depicting a graded p ?n +junction with majority of the depletion region on the p-doped segment of the nanowire,and (d)corresponding space-charge region in the device.

Figure3c shows the spatial photocurrent along the length of a NW p ?n junction device of Figure 3a for a series of applied dc biases from ?1V reverse bias to +1V forward bias.The photocurrent does not show a strong bias dependence as the reverse bias is increased from 0to ?1V.This behavior is consistent with the fact that the photocurrent is limited by the optical generation rate of the minority carriers in the neutral region and not by the depletion region width or electric field strength.The carriers generated in the depletion region and within one diffusion length will get swept across the junction due to the built-in electric field independent of the external applied reverse bias and the strength of the electric field.However,as the bias is increased in the forward direction from 0to 1V,the magnitude of the photocurrent decreases sharply implying that the current in forward bias is dominated by recombination in the depletion region.This conclusion is further supported by the greater than one ideality factor of the p ?n diode.This enhanced recombination in forward bias decreases the overall photocurrent yield since minority carrier electrons (holes)will have a reduced probability of reaching the n-type (p-type)material which in turn decreases their

contribution to the photocurrent.Furthermore,the rate of exponential decay of the photocurrent and thus the minority carrier diffusion lengths do not vary appreciably in the range of applied dc bias (Figure 3d),consistent with minority carrier diffusion limited transport within the neutral region of the NW with a negligible drift component.We also observe that the position of the photocurrent peak shifts toward the p-doped segment as we tune the bias from forward to reverse direction.With increasing reverse bias,the geometric center of the depletion region moves further into the more lightly doped p region of the NW.This observation is consistent with the asymmetric p ?n +type junction for these NWs as determined by four point measurements and discussed earlier.

The results presented here for L e eff of ～1.0μm can be confirmed by comparing to an estimated L eff determined from minority carrier lifetime measurements by Jung et al.for Si NW p ?n junctions grown under the same conditions.31From Jung et al.’s directly measured effective minority carrier lifetimes of ～40?70ns for NW diameters ～70nm,we estimate a minority carrier diffusion length for electrons of ～4μm,32which is only moderately larger than observed

here.

Figure 3.(a)Semi log plot of the photocurrent profile for the Si nanowire p ?n junction measured at 0V.The red and blue solid lines are exponential fits to obtain the effective electron and hole diffusion lengths.The dotted vertical lines mark the inner edge of the contacts.(b)Semi log plot of the photocurrent profile for a Schottky barrier device contacting an n-type segment measured at 0V.The dotted line represents the inner edge of the Schottky contact.(c)Photocurrent plotted as a function of position along nanowire for the applied dc bias stepped from +1V forward bias to ?1V reverse bias.The magnitude of photocurrent is represented by the rainbow scale.The dotted green line shows the photocurrent profile measured at 0V.(d)Effective minority carrier diffusion lengths plotted as a function of applied dc bias for the p ?n junction measurements of (c).

The observed electron and hole effective minority carrier

diffusion lengths of L e eff ≈1.0μm and L h eff

≈0.66μm are much shorter compared to those observed in bulk Si with similar doping concentrations (i.e.,L e ≈300μm and L h ≈30μm)33,34due to the close proximity of the surface for NWs.However our measured NW values are an order of magnitude longer than those previously reported in NWs of similar diameter (<100nm).23,24To compare our results with all prior reported minority carrier diffusion lengths on NWs of varying diameters and dopant concentrations,we have extracted the surface recombination velocity as a metric.The surface recombination velocity on NWs was first proposed by Allen et al.in the Supporting Information of ref 23for relating the effective recombination lifetime,τeff ,to the bulk minority carrier lifetime,τb ,where for a nanowire diameter d NW and surface recombination velocity S the effective recombination lifetime is given by

τ=τ+S d 1/1/4/eff b NW

(1)

Substituting for τeff in eq 1and ignoring the negligible contribution of the bulk recombination term (i.e.,for τb ?τeff ),we get an expression for S in terms of L eff as S =d NW D /(4(L eff )2).The relative magnitude of the surface recombination velocity provides insight into the relative magnitude of the surface state density by relating the recombination efficiency of photogenerated electron (or holes)with the surface states.Table 1compares reported L eff and evaluated S values from the present and past published reports of nanowires and bulk crystalline silicon.For the small diameter NWs measured in refs 23and 24,the minority carrier diffusion length and the surface recombination velocity are similar prior to surface passivation.However,after surface passivation with a-Si (ref 24),a minority carrier diffusion length was reported as being 7?11times greater,whereas the surface recombination velocity decreased by 2orders of magnitude which implies that the surface state density was dramatically reduced.In our case,we observe much longer minority carrier diffusion lengths and reduced surface recombination velocities (6×102to 4×103cm/s)without any additional surface passivation,suggesting that our as-grown NWs are well passivated and have a low surface state density.Also,in comparison to the as-grown large diameter (1?2μm)Si microwires which were grown with SiCl 4source gases at higher temperatures (～1000°C)and with both Au and Cu as the growth catalysts,the effective minority carrier diffusion length is comparable to our values for much smaller diameter Si NWs before passivation.In fact the surface recombination velocity in ref 26is 2orders of magnitude larger than that of our Si NWs prior to passivation.The above results suggest that the density of surface states in our in situ doped p ?n junction NWs grown in a cold wall reactor is much reduced from that of previous reports where in all cases the NWs were grown in hot-wall reactor systems.The improved surface recombination velocities found here may result from reduced precursor reactions or deposition on the NW sidewalls resulting in a lower density of recombination centers at the NW surface.The submission that the Si NWs in the present study,which were synthesized with a cold-wall CVD reactor,have a reduced surface state density is further supported by measuring the effective minority carrier diffusion lengths with and without the presence of a background illumination.Results for the effective minority carrier diffusion lengths of a NW at 0V are shown in Figure 4for the conditions with and without broad area dc background illumination in the low-injection limit provided

by

T a b l e 1.C o m p a r i s o n o f E x p e r i m e n t a l R e s u l t s o n M e a s u r i n g t h e M i n o r i t y C a r r i e r D i f f u s i o n L e n g t h i n S i N a n o w i r e s

h A u /p -S i /A u S c h o t t k y A u 1×1018

～25?100(L e )～3×105

～250?400(L e )

～4×10

3

24N i /n -S i /p -S i /N i a x i a l p n j u n c t i o n s

A u

2×1016

2×10

18

～660(L h )～4×103

o u r w o r k

～1000(L e )～6×102

A l /n -S i /A l S c h o t t k y A u l a r g e d m i c r o w i r e s (1?2μm )

S P C M 1×1017

～2000(L h )～1×103

25A l /p -S i /n -S i /A l e x s i t u d o p e d p n j u n c t i o n s C u 1×1017

1×1017

≤500(L e )>4×105

≥30000(L e )

<10

26c -S i

b u l k S i

E B I C

2×1016

2×1018

～50000(L h )<1

33a n d 34

～250000(L e )

a separate laser of λ=900nm.A similar experiment was performed by Putnam et al 27on large diameter (d ≈2.0μm)Si microwires where effective diffusion lengths increased from ≤0.7μm to ～10μm in the presence of broad area illumination.This increase was attributed to increased surface passivation due to the filling of surface states by the photogenerated carriers created by the background light.In contrast,we find that there is very little change in the effective minority carrier diffusion lengths measured with and without the background light.This result supports the view that the present as-grown NWs have significantly fewer surface states.In addition,in another report 35it was determined using TEM that InAs NWs grown with a cold-wall system had high crystalline quality and much reduced defect density as compared to InAs NWs grown with a hot-wall reactor.However,to truly understand the variations in surface recombination for NWs due to subtle variations in the growth conditions such as for different types of reactors (hot wall versus cold wall),growth temperatures,and partial pressures,a quantitative comparison across NWs grown under different conditions from different laboratories is required and beyond the scope of this study.

In summary,surface recombination provides the major limitation for achieving long effective minority carrier diffusion lengths in nanowires and thus control of surface quality is of key importance in achieving efficient photocurrent response essential for solar cell applications.We have reported here effective minority carrier diffusion lengths for single crystalline Si nanowire p ?n junctions which are an order of magnitude larger than previously reported for as-grown small diameter nanowires (<100nm).These devices show a high internal quantum efficiency of ～50%,consistent with low surface recombination velocities compared to those found in previous as-grown nanowire and microwire studies.These observations of long effective minority carrier diffusion lengths in our nanowires suggests that the surface state density is lower in comparison with previous reports and are promising for more efficient future photovoltaic device applications.

ASSOCIATED CONTENT

*

Supporting Information Control experiments to verify the observed large values of L eff .This material is available free of charge via the Internet at https://www.wendangku.net/doc/fc12852641.html,.

■

AUTHOR INFORMATION

Corresponding Author

*E-mail:amohite@https://www.wendangku.net/doc/fc12852641.html,;picraux@https://www.wendangku.net/doc/fc12852641.html,;htoon@https://www.wendangku.net/doc/fc12852641.html,.

Present Addresses

∥

Environmental Molecular Science Laboratory,Pacific North-west National Laboratory,Richland,WA 99352.⊥

School of Electrical and Computer Engineering,Georgia Institute of Technology,Atlanta,GA 30332.

Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS

We thank John Baldwin from the Center for Integrated Nanotechnologies (CINT)at Los Alamos National Laboratory (LANL)for his help in preparing the growth substrates needed for the nanowire synthesis.This work was performed,in part,at CINT,a U.S.Department of Energy,Office of Basic Energy Sciences user facility.The research was funded in part by the Laboratory Directed Research and Development Program and by CINT at LANL,an affirmative action equal opportunity employer operated by Los Alamos National Security,LLC,for the National Nuclear Security Administration of the U.S.Department of Energy under contract DE-AC52-06NA25396.

■

REFERENCES

(1)Shah,A.;Torres,P.;Tscharner,R.;Wyrsch,N.;Keppner,H.Science 1999,285(5428),692?698.

(2)Brendel,R.Introduction,in Thin-Film Crystalline Silicon Solar Cells:Physics and Technology ;Wiley:New York,2003.(3)Lewis,N.S.Science 2007,315(5813),798?801.

(4)Boettcher,S.W.;Spurgeon,J.M.;Putnam,M.C.;Warren,E.L.;Turner-Evans,D.B.;Kelzenberg,M.D.;Maiolo,J.R.;Atwater,H.A.;Lewis,N.S.Science 2010,327(5962),185?187.

(5)Goodey,A.P.;Eichfeld,S.M.;Lew,K.K.;Redwing,J.M.;Mallouk,T.E.J.Am.Chem.Soc.2007,129(41),12344.

(6)Gunawan,O.;Wang,K.;Fallahazad,B.;Zhang,Y.;Tutuc,E.;Guha,S.Prog.Photovoltaics 2011,19(3),307?312.

(7)Kendrick,C.E.;Yoon,H.P.;Yuwen,Y.A.;Barber,G.D.;Shen,H.;Mallouk,T.E.;Dickey,E.C.;Mayer,T.S.;Redwing,J.M.Appl.Phys.Lett.2010,97(14),143108.

(8)Maiolo,J.R.;Kayes,B.M.;Filler,M.A.;Putnam,M.C.;Kelzenberg,M.D.;Atwater,H.A.;Lewis,N.S.J.Am.Chem.Soc.2007,129(41),12346.

(9)Putnam,M.C.;Boettcher,S.W.;Kelzenberg,M.D.;Turner-Evans,D.B.;Spurgeon,J.M.;Warren,E.L.;Briggs,R.M.;Lewis,N.S.;Atwater,H.A.Energy Envron.Sci.2010,3(8),1037?1041.

(10)Stelzner,T.;Pietsch,M.;Andra,G.;Falk, F.;Ose, E.;Christiansen,S.,Nanotechnology 2008,19(29).

(11)Tsakalakos,L.;Balch,J.;Fronheiser,J.;Korevaar,B.A.;Sulima,O.;Rand,J.,Appl.Phys.Lett.2007,91(23).

(12)Picraux,S.T.;Yoo,J.;Campbell,I.H.;Dayeh,S.A.;Perea,D.E.Semiconductor Nanowires for Solar Cells.In Semiconductor nanostruc-tures for optoelectronic devices ;Springer:New York,2011.

(13)Wagner,R.S.;Ellis,W.C.,Appl.Phys.Lett.1964,4(5),89-&.(14)Cui,Y.;Lauhon,L.J.;Gudiksen,M.S.;Wang,J.;Lieber,C.M.,Diameter-controlled synthesis of single-crystal silicon nanowires .AIP:2001;Vol.78,p 2214?2216.

Figure 4.Semi log plots of photocurrent profiles of a Si nanowire p ?n junction measured at 0V without background illumination (open circles)and with background illumination (open triangles).The fit to the exponential decay of the photocurrent gives the effective electron and hole diffusion lengths.

(15)Kelzenberg,M.D.;Boettcher,S.W.;Petykiewicz,J.A.;Turner-Evans,D.B.;Putnam,M.C.;Warren,E.L.;Spurgeon,J.M.;Briggs,R. M.;Lewis,N.S.;Atwater,H.A.Nat.Mater.2010,9(3),239?244.

(16)Kempa,T.J.;Tian,B.;Kim,D.R.;Hu,J.;Zheng,X.;Lieber,C. M.Nano Lett.2008,8(10),3456?3460.

(17)Gudiksen,M.S.;Lauhon,L.J.;Wang,J.;Smith,D.C.;Lieber,

C.M.Nature2002,415(6872),617?620.

(18)Tian,B.;Zheng,X.;Kempa,T.J.;Fang,Y.;Yu,N.;Yu,G.; Huang,J.;Lieber,C.M.Nature2007,449(7164),885?889. (19)Cui,Y.;Duan,X.;Hu,J.;Lieber,C.M.J.Phys.Chem.B2000, 104(22),5213?5216.

(20)Kelzenberg,M.D.;Putnam,M.C.;Turner-Evans,D.B.;Lewis, N.S.;Atwater,H.A.In Predicted efficiency of Si wire array solar cells; Photovoltaic Specialists Conference(PVSC),200934th IEEE;2009; pp001948-001953.

(21)Bullis,W.M.Solid State Electron.1966,9(2),143.

(22)Sze,S.M.Physics of Semiconductor devices;Wiley-Interscience: New York,1969;p xiv?812.

(23)Allen,J.E.;Hemesath,E.R.;Perea,D.E.;Lensch-Falk,J.L.;Li, Z.Y.;Yin,F.;Gass,M.H.;Wang,P.;Bleloch,A.L.;Palmer,R.E.; Lauhon,L.J.Nat.Nanotechnol.2008,3(3),168?173.

(24)Dan,Y.;Seo,K.;Takei,K.;Meza,J.H.;Javey,A.;Crozier,K.B. Nano Lett.2011,11(6),2527?2532.

(25)Kelzenberg,M.D.;Turner-Evans,D.B.;Kayes,B.M.;Filler,M.

A.;Putnam,M.C.;Lewis,N.S.;Atwater,H.A.Nano Lett.2008,8(2), 710?714.

(26)Kelzenberg,M. D.;Turner-Evans, D. B.;Putnam,M. C.; Boettcher,S.W.;Briggs,R.M.;Baek,J.Y.;Lewis,N.S.;Atwater,H.A. Energy Envron.Sci.2011,4(3),866?871.

(27)Putnam,M. C.;Turner-Evans, D. B.;Kelzenberg,M. D.; Boettcher,S.W.;Lewis,N.S.;Atwater,H.A.Appl.Phys.Lett.2009, 95,16.

(28)Picraux,S.T.;D.,S.;Manandhar,P.;Perea,D.E.;Choi,S.G.J. Mater.2010,62(4),35.

(29)Gu,Y.;Romankiewicz,J.P.;David,J.K.;Lensch,J.L.;Lauhon, L.J.Nano Lett.2006,6(5),948?952.

(30)It should be noted that in all of the devices Ni was used to make contacts to the NWs.Ni typically is known to form Schottky contact to n-type Si NWs,but the Schottky barrier height is strongly dependent on the local doping concentration.In our study we found that out of20two terminal devices fabricated using Ni as the metal contact only three devices showed lower photocurrent at both the p and n contacts(ohmic)while for other devices the n segment formed

a Schottky contact to Ni.

(31)Jung,Y.;Vacic,A.;Perea,D.E.;Picraux,S.T.;Reed,M.A.Adv. Mater.2011.

(32)A diffusion constant of D e≈3.38cm2/s and D h≈9.8cm2/s was calculated using bulk mobilities for Si at the doping concentrations estimated from resistivity measurements.

(33)del Alamo,J.A.;Swanson,R.M.Solid State Electron.1987,30 (11),1127?1136.

(34)Tyagi,M.S.;Van Overstraeten,R.Solid State Electron.1983,26 (6),577?597.

(35)Schroer,M.D.;Xu,S.Y.;Bergman,A.M.;Petta,J.R.Rev.Sci.

Instrum.2010,81(2).