In Situ Characterization of Nanowire Dimensions and Growth

Mar 25, 2015 - Christian Camus,. ‡. Lars Samuelson,. † and Magnus T. Borgström. †. †. Division of Solid State Physics, Lund University, 22100...
0 downloads 0 Views 820KB Size
Subscriber access provided by NEW YORK MED COLL

Communication

In Situ Characterization of Nanowire Dimensions and Growth Dynamics by Optical Reflectance Magnus Heurlin, Nicklas Anttu, Christian Camus, Lars Samuelson, and Magnus T Borgstrom Nano Lett., Just Accepted Manuscript • Publication Date (Web): 25 Mar 2015 Downloaded from http://pubs.acs.org on March 26, 2015

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

Nano Letters is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 20

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Nano Letters

In Situ Characterization of Nanowire Dimensions and Growth Dynamics by Optical Reflectance Magnus Heurlin1,† *, Nicklas Anttu1,†, Christian Camus2, Lars Samuelson1 and Magnus T. Borgström1, †M. H. and N. A. contributed equally to this work 1

Division of Solid State Physics, Lund University, 22100 Lund, Sweden

2

LayTec AG, Seesener Str. 10-13, 10709 Berlin, Germany

KEYWORDS: Optical metrology, in-situ measurement, nanowire, MOVPE, nanoimprint lithography, indium phosphide

Abstract

Optical reflectometry is commonly used as an accurate and non-invasive characterization tool when growing planar semiconductor layers. However, thin-film analysis schemes cannot be directly applied to nanowire systems due to their complex optical response. Here, we report on reliable in situ characterization of nanowire growth with high accuracy using optical reflectance spectra for analysis. The method makes it possible to determine the nanowire length, diameter, and growth rate in situ, in real time, with high resolution. We demonstrate the method’s versatility by using the optical reflectance data for determining nanowire dimensions on both

ACS Paragon Plus Environment

1

Nano Letters

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 20

particle assisted and selective-area grown nanowires. To indicate the full potential of in situ characterization of nanowire synthesis we evaluate the growth dynamics of InP nanowires in the presence of the p-type dopant precursor diethylzinc. We observe that the growth rate is strongly affected by the diethylzinc. At low diethylzinc flows the growth rate decreases monotonously while higher flows leads to an initially increasing growth rate. From these in situ characterization data, we conclude that the surface migration length of adatom species is affected strongly by the addition of diethylzinc. We believe that this characterization method will become a standard tool for in situ growth monitoring and aid in elucidating the complex growth dynamics often exhibited during nanowire growth.

In situ monitoring of nanowire growth is highly desirable since these nanostructures will play a key role in future semiconductor devices such as light-emitting diodes,1 photovoltaic cells,2 transistors3 and medical sensors4. In situ studies of crystal growth have been important for the development of reliable fabrication processes and understanding how crystal surfaces evolve. With techniques such as in situ scanning tunneling microscopy, reflection high energy electron diffraction (RHEED) and optical reflectometry it is possible to monitor adatom movement on a crystal surface,5 study how the surface reconstruction changes,6 and determine growth rate and layer composition7. Monitoring of nanowire growth in situ (but not under typical growth conditions in real systems) has also played a key role in developing our understanding of nanowire growth.8 For example, several studies have been performed in specially developed transmission electron microscopes where observations of the dynamical particle-nanowire interface during growth have shown that it can proceed in an oscillatory manner9 where the frequency can depend on twin defect formation10. Furthermore, studies of nanowire growth with mass spectrometry11 and RHEED12 inside molecular beam epitaxy (MBE) systems have been

ACS Paragon Plus Environment

2

Page 3 of 20

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Nano Letters

used to determine nucleation conditions and crystal structure formation in situ. Infrared spectroscopy has also recently been used to extract information on the surface chemistry during nanowire growth.13 In common for these in situ studies of nanowire growth is the use of low growth pressures not realistic inside a metal-organic vapor phase epitaxy (MOVPE) system, which is the most prevalent technique for large volume fabrication of high performance III-V semiconductor devices. Furthermore, all of the above mentioned techniques apply to either very small volumes, can only determine the time of nucleation, extract structural information from xray diffraction or study the nanowire surface chemistry. However, there is a strong need to determine the nanowire dimensions in large-area samples in situ to take full control over the growth process, for example in synthesis of nanowire arrays used in opto-electronic applications.2, 14-16 Optical reflectometry is commonly used for characterizing large-area samples when growing planar semiconductor layers in an MOVPE.17 However, such thin-film analysis schemes cannot be applied directly to nanowire systems due to their complex optical response. Ex-situ reflectometry measurements at room temperature and atmospheric pressure has successfully been used in combination with electrodynamic modelling to evaluate nanowire dimensions,18 but due to the computationally intensive calculations this is not suitable for real time measurements. Previously, in situ single wavelength optical reflectometry was used to determine the incubation time for the onset of growth and the volume fill fraction of Si nanowires during chemical vapor deposition, giving a measure of the nanowire growth rate.19 In order to establish optical in situ metrology as a generic tool for analyzing nanowire length with high precision, a careful estimation of the effective refractive index is needed, not assuming a simple linear super position between the amount of vapor and nanowire material.19, 20 To asses the full nannowire properties

ACS Paragon Plus Environment

3

Nano Letters

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 4 of 20

in situ, being able to characterize the nanowire diameter is also necessary, for which no solution currently exsists. Moreover, such optical metrology would benefit from high temporal resolution, enabling access to the complex and often non-linear nanowire growth dynamics. Here, we report the development of a method for in situ characterization of nanowire growth rate, length, and diameter by analysis of multiple wavelength optical reflectance spectra which have a high temporal resolution. This method does not rely on computationally intensive modelling, making it suitable for online real time monitoring. The nanowire length is determined by observing the constructive and destructive interference which occurs between light reflected at the nanowire top and at the nanowire-substrate interface. By experimentally measuring an effective refractive index for a given nanowire array, we can convert this interference signal directly into a nanowire length. The nanowire diameter is in turn determined by tracking wavelength dependent absorption resonances21, 22 which are independent of the nanowire length and the array pitch. Since the analysis to obtain the nanowire length and diameter only uses an effective refractive index and a linear equation, which relates the absorption resonance to a certain diameter, as input, it can be applied by anyone with an optical reflectance setup connected to a growth reactor. To show the versatility of these in situ measurements we first use it to measure the length of particle assisted grown InP nanowires. The nanowires are grown (see Supporting Information Figure S1 for scanning electron microscope (SEM) images) from nanoimprint lithography (NIL) defined Au particle arrays (400 nm pitch) inside an Aixtron® 200/4 MOVPE equipped with a LayTec® EpiR DA UV optical reflectometry system. This allows us to measure one wavelength spectrum per second in the 400 - 800 nm wavelength range at normal incidence. An example of the reflectance spectra obtained during an InP nanowire growth run is shown in Figure 1a where

ACS Paragon Plus Environment

4

Page 5 of 20

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Nano Letters

a set of fringes are visible, as alternating dark and bright bands. These fringes originate from alternating constructive and destructive interference between light reflected at the nanowire top and the nanowire-substrate interface. In order to convert this interference information into nanowire length, we note that the first fringe (a minimum) corresponds to an optical path length difference of half a wavelength.a Each subsequent maxima or minima correspond, in turn, to an additional path length difference of half a wavelength. Therefore, we can determine the nanowire length from a measured minimum or maximum for a given measured wavelength according to: ‫=ܮ‬

௠ఒ ସ௡

.

(1)

Here, L denotes the nanowire length, λ is the wavelength of the light, n is the refractive index of the nanowire-array layer, and m denotes the minima and maxima that show up consecutively as the growth time progresses (m = 1, 3, 5, … for minima and m = 2, 4, 6, … for maxima). Knowledge of the refractive index n of the nanowire layer is necessary for using this approach. Values for n can be obtained in several ways. One method is to approximate an effective refractive index based on the averaged volume of InP and vapor. However such an effectivemedium theory is not easy to implement to yield reliable results,20, 21 since the light couples into specific optical modes of the nanowire array. Another method is to extract the refractive index from the modeled optical response of the same nanowire array. It would however be difficult to perform such computationally intensive electromagnetic modeling for all the relevant nanowire array combinations that can be encountered. Furthermore, some parts of the nanowire system

a

We note that the light that is reflected at the top interface between the air superstrate and the nanowire array is incident toward an optically denser medium. Therefore, we expect an additional phase shift of π due to this reflection. However, also the light that is reflected at the nanowire-array/substrate interface is incident toward a denser medium which gives the same phase shift. Therefore the phase shift does not have to be taken into account.

ACS Paragon Plus Environment

5

Nano Letters

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 6 of 20

might be hard to model accurately, such as the catalyst particle due to for example unknown alloy composition during the growth.

Figure 1. (a) Reflectance spectra measured during nanowire growth. The growth run starts with an annealing period (not shown in its entirety) before nanowire growth is initiated at t = 0 s. The total nanowire growth time is 1200 s. (b) Effective refractive index as function of λ extracted from the mean of 6 different nanowire growth runs. Between 550 nm and 700 nm a resonance exists in the absorption spectra of the nanowires which prohibit us from extracting data due to strong absorption. The strong absorption results in a low intensity reflectance signal with weak oscillations since the light that is reflected at the nanowiresubstrate interface is absorbed to a high degree. Instead, we chose to perform a series of calibration runs using the same imprint pattern as during the actual final nanowire fabrication. From these calibration runs, we calculated an average refractive index (neff which we used for n in Eq. 1) (see the Supporting Information for details). This averaging reduces the effect of statistical fluctuations that could occur in a single calibration run and increases the reliability of the obtained neff. The calibration was performed by first measuring the nanowire length from the calibration runs with SEM. Next, by rearranging

ACS Paragon Plus Environment

6

Page 7 of 20

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Nano Letters

Eq. (1), neff could be extracted for each wavelength (see Figure 1b). Between 550 and 700 nm, a diameter dependent absorption resonance exists for this specific nanowire diameter, which prohibits the extraction of neff through the interference maxima and minima since the signal is too weak. By using the analysis outlined in the previous paragraph, we can determine the nanowire length during the growth run presented in Figure 1a. For this purpose we used a wavelength range containing 101 different wavelengths since each minimum or maximum for adjacent wavelengths shows up at slightly different growth time (see the Supporting Information for details). We selected the wavelength interval between 420 and 520 nm since here the interference oscillations are strong, and we avoid the absorption resonance at 550 < λ < 700 nm. The obtained nanowire length as a function of growth time is shown in Figure 2a together with the nanowire length, as measured by SEM, from four growth runs with varying growth time. When we compared the final length obtained from SEM with that obtained from in situ measurements for twelve different samples, we found an agreement within 3% for the final nanowire length with a standard deviation of 2 percentage points. The temporal resolution achieved depends on the number of wavelengths used and the nanowire growth rate. In our case, we detect on average at least one minimum or maximum per second which results in an almost continuous monitoring of the nanowire length. Since the temporal resolution is high, we were able to determine the growth rate from the derivative of the extracted relationship between the nanowire length and the growth time (Figure 2b). In the measurements we observe a variation in the extracted growth rate for short nanowire lengths (below 500 nm). We believe that this originates from a slight change in the wavelength dependence of neff at short nanowire lengths (typically when the nanowires are shorter than the wavelength of light used).

ACS Paragon Plus Environment

7

Nano Letters

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 8 of 20

Figure 2. (a) Nanowire length as function of time extracted from a wavelength interval between λ = 420 - 520 nm. The red data points indicate the nanowire length measured by SEM from four growth runs with different growth time. The standard deviations from the SEM measurements were approximately 30 nm for the 30 nanowires measured on each sample and is indicated by the error bars. (b) Nanowire growth rate as function of nanowire length (top) and growth time (bottom) as extracted from the data in (a). To determine the nanowire diameter, we can track optical resonances in the individual nanowires which are strongly diameter dependent.21 To demonstrate this, we grew InP nanowires with a selective area growth mechanism since core-shell structures can be grown reliably using this technique.23 The sites where nanowire growth was initiated were defined by the use of NIL but with a 1 µm pitch instead of the denser pattern used previously. This change was made in order to show the versatility of the method for varying pitch. During the growth run, both axial and radial growth was performed and the resulting reflectance data is shown in Figure 3a.

ACS Paragon Plus Environment

8

Page 9 of 20

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Nano Letters

Figure 3. (a) Reflectance of a nanowire array measured during growth with selective area in an array with a pitch of 1 µm. Black dashed (dash-dot) line indicate start and stop of nanowire core (shell) growth. White dashed line indicates the absorption resonance giving a minimum in reflectance. (b) Nanowire length as a function of time extracted from a wavelength span between 430 to 480 nm for the nanowire growth between 0 and 1200 s [indicated by the dashed lines in (a)]. (c) Nanowire diameter as function of growth time. Shell growth is initiated at a time of 1530 s and stopped at 2730 s [indicated by dash-dotted lines in (a)]. The red points in (b) and (c) show the nanowire length and diameter measured post growth by SEM. In (b) the SEM measurement was performed on a different sample grown with the same reactor conditions but without the shell growth step. The error bars in (b) and (c) indicate the standard deviation from 30 measured nanowires. For calculating the nanowire length as a function of time (Figure 3b), we found that the coupling into the nanowires was weak, due to the large pitch, such that neff = 1 for all wavelengths could be used (see the Supporting Information for details on extracting the length of less dense arrays). During the nanowire shell growth step, the nanowire diameter was in turn determined (Figure 3c) by following the shift of an absorption resonance which shows up as a dip in the reflectance measurements (Supporting Information Figures S5 and S6). Specifically, we monitor resonant absorption through the HE12 optical mode of the nanowires.21 The HE12 resonance is visible as the dark blue band that starts at λ ≈ 450 nm for a growth time of 1500 s

ACS Paragon Plus Environment

9

Nano Letters

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 10 of 20

and extends to λ ≈ 650 nm for 2700 s in Figure 3a and is indicated by the white dashed line. Notably, both the diameter and the length we determine at the end of the growth time with this in situ method are within one standard deviation of the values determined with SEM after growth (Figures 3b and 3c). Regarding the determination of the nanowire diameter, we note that with the HE12 resonance it is possible to track the nanowire diameter from 163 to 381 nm within the measured wavelength interval. For thinner nanowires, the HE11 resonance can be used down to a diameter of 53 nm. However, the fringes in the reflectance signal will decrease in amplitude with the nanowire diameter due to decreasing reflection at the top of the nanowires. Therefore, we expect the signal to noise ratio to set the limit for the minimum nanowire diameter that can be monitored. Notice that the HE11 resonance is visible close to 800 nm in Figure 3a during the core nanowire growth (0 s < t < 1200 s), and it is responsible for the strong absorption between 550 and 700 nm in Figure 1a. Next, to demonstrate the potential of the in situ method, we investigated the effect of adding diethylzinc (DEZn) to the vapor flow during growth of Au nucleated InP nanowires. DEZn is commonly used as a p-type dopant for InP nanowires.24 Therefore, any effect on the growth dynamics using DEZn is highly important for energy relevant device applications requiring ptype doping, such as solar cells and LEDs, and for future IC structures which utilize p-type channels in for example CMOS structures. We employed the imprint pattern defined by 400 nm pitch and approximately 130 nm Au alloy particle size after annealing. We also note that for moderate degenerate doping, the refractive index of InP is only slightly modified (below 1% for a doping level of 3*1018 cm-3, for the wavelength region used here25) and will thus not influence our results in a noticeable manner. In Figure 4a we show the growth rate vs. nanowire length for

ACS Paragon Plus Environment

10

Page 11 of 20

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Nano Letters

six different DEZn molar fractions (χDEZn). We can first observe that as the amount of DEZn increases, the final nanowire length increases for the fixed growth time of 20 min (see Supporting Information Figure S7 for SEM images). The nanowire length increases from approximately 2 µm at the lowest χDEZn to 3.5 µm for the highest χDEZn, without adding any additional growth material, i.e., indium (In) or phosphorous (P). To understand this phenomenon, we assess the data from the full DEZn doping series and note that for the three highest χDEZn the growth rate has distinctly different behavior as compared to lower χDEZn. For low χDEZn the growth rate is initially constant until it reduces after the nanowires reach a length of approximately 700 nm. In stark contrast, for the three highest χDEZn the growth rate initially increases as a function of time (or equivalently, nanowire length) and decreases only after reaching a length of approximately 850 nm, 1 µm and 1.5 µm for χDEZn = 3.0*10-5, χDEZn = 5.0*10-5, and χDEZn = 6.0*10-5 respectively. This dependency indicates that the increased growth rate when using higher χDEZn is related to the surface migration length of adatom species being affected by the addition of DEZn. The phenomena can then be understood from a model developed by Borg et al,26 which relates the available collection area (which will vary with the length of the nanowire and the migration length) of the rate limiting growth species, to changes in the nanowire growth rate. Next, we resolved whether the reason for the apparent longer surface migration length of adatoms is due to a change from a mixed crystal structure of zinc blende and wurtzite to pure zinc blende, induced by the Zn doping.27 Also for the investigation of this advanced topic, the presented in situ method is ideally suited. We designed an experiment where the DEZn flow (χDEZn = 6.0*10-5) was periodically turned on and off during nanowire growth. In Figure 4b, we show the nanowire length as a function of time for such a barcoded structure. We note that

ACS Paragon Plus Environment

11

Nano Letters

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 12 of 20

growth in principle stops when turning the DEZn flow off, and reinitiates when turning the flow on. Thus, the effect cannot be explained by a change in crystal structure induced by Zn doping,27 since the grown surface facets should contribute by diffusion to growth of the following segment. Instead, we argue that an accumulation layer of Zn is formed on the substrate and nanowire side facets under sufficiently high DEZn flows in accordance with the results of Haruki et al,28 who observed that excess Zn can occupy surface sites where In would normally incorporate and inhibit the incorporation of In. Thus, more In can migrate to the Au particle where it can incorporate at the nanowire growth front, increasing the nanowire growth rate. The fact that the growth rate drops when the Zn flow is turned off in the barcoded structure indicates that the Zn accumulation layer desorbs during this step due to its high vapor pressure.29 This again reduces the diffusion length when growing the intrinsic segments and decreases the nanowire growth rate. Once the nanowire length is longer than the surface migration length, the growth rate should reach a constant value. In our case it continuously drops as the nanowires grow longer for all investigated samples. This dependency indicates that the main part of In species arriving at the growth front originates from surface diffusion on the substrate surface rather than from direct impact on the Au particle and the nanowire side walls.

ACS Paragon Plus Environment

12

Page 13 of 20

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Nano Letters

Figure 4. (a) Growth rate as a function of nanowire length for 6 different DEZn flows. (b) Nanowire length vs. time for a barcode structure with alternating DEZn doped (χDEZn = 6.0*10-5) and intrinsic segments. The growth of each segment is 120 s long except for a 30 s intrinsic nucleation step and the final segment which is intrinsic and grown for 90 s. In later stages of growth after 870 s, gaps in the curve indicate that no minima or maxima are recorded and the growth rate is thus close to zero. We expect this non-invasive and purely optical in situ analysis method to be used as a standardized tool for monitoring nanowire growth. The demonstrated ability to monitor the nanowire diameter and length in arrays of varying geometry indicates the general applicability of the technique. Since the analysis is not dependent on a specific choice for the nanowire material, we expect the method to be generic and applicable also to other commonly grown materials, such as the opto-electronically important semiconductors GaAs and InAs. In addition to being used as an efficient process monitoring tool, the method opens up for fundamental in situ studies under normal growth conditions inside for example MOVPE or MBE reactors. We demonstrated such an in situ study here by deciphering the complex growth dynamics of Zn doped InP nanowires, which without the presented in-situ technique would have to be performed with lower resolution and considerably more time effort. We estimate that a similar (but not as well resolved) investigation using conventional techniques, such as SEM, would require at least 10 times more growth experiments. Additionally, accurate determination of the nanowire length, required for the barcoded structure, would be difficult to achieve with SEM and not be able to provide the same amount of data as presented here carrying information on the full nanowire growth process. Our analysis relies on the effective refractive index neff of the nanowire array. Therefore, we intend to study in detail the dependence of neff on the array pitch as well as the diameter and

ACS Paragon Plus Environment

13

Nano Letters

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 14 of 20

material of the nanowires. Such understanding will enhance the easy application of the method to a large class of nanowire systems. Currently our measurements indicate that a neff of 1 can be used for pitches larger than 1 µm which simplifies the analysis for a large range of nanowire arrays. Furthermore, the study of shorter nanowires (below 500 nm in length) in dense arrays is an interesting future direction, since short nanowires could show a different type of interaction with light than long nanowires. The optical response of long nanowires can often be described with an effective refractive index as demonstrated here, but for nanowires of sub-wavelength length, the concept of an effective refractive index must be studied in detail since light scattering at the top and the bottom end facet of the nanowire can dominate the optical response. Limitations with respect to the length of the nanowire will be encountered if the absorption of light in the nanowire array is too high, similar to synthesis and characterization of planar layers. This limitation can be circumvented by selecting a wavelength interval for the measurement where the material under investigation doesn’t absorb strongly. The presented analysis method can already now be used to monitor and define the length of different doped segments inside nanowire-based solar cell and LED structures. Such control will lead to increased process reliability, more consistent device performance, and aid in establishing nanowires as a mature technology. Methods Nanowire growth was performed from arrays defined using nanoimprint lithography in an Aixtron 200/4 MOVPE with a total flow of 13 l/min at a working pressure of 100 mbar. For particle assisted growth the Au catalyst particles were defined on 2” InP (111)B wafers using evaporation and lift-off. The selective area growth mask was defined on 2” InP (111)A wafers by dry etching of 20 nm SiNx deposited on the InP surface, using the imprint polymer pattern as an

ACS Paragon Plus Environment

14

Page 15 of 20

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Nano Letters

etch mask. Particle assisted growth was performed at 440 °C with χTMIn = 6.5*10-5 (6.5*10-5), χPH3 = 6.9*10-3 (2.3*10-3) and χHCl = 6.2*10-5 (0) during growth (nucleation). The nanowires analyzed in Figures 1a and 2 were intrinsic while the nanowires analyzed in Figure 1b and Figure 4 were grown with a varying amount of DEZn. Axial selective area growth was performed at 700 °C with χTMIn = 1.1*10-5 and χPH3 = 1.9*10-3 while the radial growth was performed at 600 °C with χTMIn = 3.2*10-5 and χPH3 = 7.7*10-3. The reflectance measurements were obtained using a LayTec® EpiR DA UV system in the 400 – 800 nm wavelength range. This system uses an XBO light source and a diode array for simultaneously obtaining the reflectance spectra of the entire wavelength range. The light was incident and collected perpendicular to the substrate through a H2 purged optical window. SEM measurements were performed post-growth to first obtain the effective refractive index and later to compare the length and diameter values obtained from the reflectance measurements with the SEM measured values. Images were acquired from three different locations on each sample and at least ten nanowires were measured in each location. ASSOCIATED CONTENT Supporting Information available: Additional SEM images and technical details of the analysis method. This material is available free of charge via the Internet at http://pubs.acs.org.

AUTHOR INFORMATION Corresponding Author *E-mail: [email protected]

ACS Paragon Plus Environment

15

Nano Letters

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 16 of 20

Author Contributions The manuscript was written through contributions of all authors. All authors have given approval to the final version of the manuscript. †M. H. and N. A. contributed equally to this work

ACKNOWLEDGMENT The authors would like to acknowledge Sebastian Lehmann, Daniel Jacobsson, and Robert Hallberg for discussions and sharing results of reflectance spectra measured on different nanowire samples. This project was performed in collaboration between the Nanometer Structure Consortium at Lund University (nmC@LU) and LayTec AG. Financial support was provided from the EU project NWs4Light under Grant No. 280773, the Swedish Research Council (VR), the Swedish Foundation for Strategic Research (SSF), The Swedish Energy Agency, the Nordic Innovation program NANORDSUN, the Knut and Alice Wallenberg Foundation (KAW) and VINNOVA. REFERENCES

1.

Huang, M. H.; Mao, S.; Feick, H.; Yan, H.; Wu, Y.; Kind, H.; Weber, E.; Russo, R.;

Yang, P. Science 2001, 292, (5523), 1897-1899. 2.

Wallentin, J.; Anttu, N.; Asoli, D.; Huffman, M.; Aberg, I.; Magnusson, M. H.; Siefer,

G.; Fuss-Kailuweit, P.; Dimroth, F.; Witzigmann, B.; Xu, H. Q.; Samuelson, L.; Deppert, K.; Borgstrom, M. T. Science 2013, 339, (6123), 1057-1060. 3.

Tomioka, K.; Yoshimura, M.; Fukui, T. Nature 2012, 488, (7410), 189-92.

ACS Paragon Plus Environment

16

Page 17 of 20

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Nano Letters

4.

Zheng, G.; Patolsky, F.; Cui, Y.; Wang, W. U.; Lieber, C. M. Nat Biotech 2005, 23, (10),

1294-1301. 5.

Tsukamoto, S.; Koguchi, N. Journal of Crystal Growth 1999, 201–202, (0), 118-123.

6.

Aspnes, D. E.; Harbison, J. P.; Studna, A. A.; Florez, L. T. Physical Review Letters 1987,

59, (15), 1687-1690. 7.

Breiland, W. G.; Killeen, K. P. Journal of Applied Physics 1995, 78, (11), 6726-6736.

8.

Wu, Y.; Yang, P. Journal of the American Chemical Society 2001, 123, (13), 3165-3166.

9.

Oh, S. H.; Chisholm, M. F.; Kauffmann, Y.; Kaplan, W. D.; Luo, W.; Rühle, M.; Scheu,

C. Science 2010, 330, (6003), 489-493. 10. Wen, C. Y.; Tersoff, J.; Hillerich, K.; Reuter, M. C.; Park, J. H.; Kodambaka, S.; Stach, E. A.; Ross, F. M. Physical Review Letters 2011, 107, (2), 025503. 11. Fernández-Garrido, S.; Zettler, J. K.; Geelhaar, L.; Brandt, O. Nano Letters 2015. 12. Tchernycheva, M.; Harmand, J. C.; Patriarche, G.; Travers, L.; Cirlin, G. E. Nanotechnology 2006, 17, (16), 4025. 13. Shin, N.; Filler, M. A. Nano Letters 2012, 12, (6), 2865-2870. 14. Svensson, C. P. T.; Mårtensson, T.; Trägårdh, J.; Larsson, C.; Rask, M.; Hessman, D.; Samuelson, L.; Ohlsson, J. Nanotechnology 2008, 19, (30), 305201. 15. Tomioka, K.; Motohisa, J.; Hara, S.; Hiruma, K.; Fukui, T. Nano Letters 2010, 10, (5), 1639-1644.

ACS Paragon Plus Environment

17

Nano Letters

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 18 of 20

16. Scofield, A. C.; Shapiro, J. N.; Lin, A.; Williams, A. D.; Wong, P.-S.; Liang, B. L.; Huffaker, D. L. Nano Letters 2011, 11, (6), 2242-2246. 17. Killeen, K. P.; Breiland, W. G. JEM 1994, 23, (2), 179-183. 18. Anttu, N.; Heurlin, M.; Borgström, M. T.; Pistol, M.-E.; Xu, H. Q.; Samuelson, L. Nano Letters 2013, 13, (6), 2662-2667. 19. Clement, T.; Ingole, S.; Ketharanathan, S.; Drucker, J.; Picraux, S. T. Applied Physics Letters 2006, 89, (16), 163125. 20. Wu, P. M.; Anttu, N.; Xu, H. Q.; Samuelson, L.; Pistol, M.-E. Nano Letters 2012, 12, (4), 1990-1995. 21. Anttu, N.; Xu, H. Q. Opt. Express 2013, 21, (S3), A558-A575. 22. Wang, B.; Leu, P. W. Opt. Lett. 2012, 37, (18), 3756-3758. 23. Premila, M.; Junichi, M.; Takashi, F. Nanotechnology 2005, 16, (12), 2903. 24. Minot, E. D.; Kelkensberg, F.; van Kouwen, M.; van Dam, J. A.; Kouwenhoven, L. P.; Zwiller, V.; Borgström, M. T.; Wunnicke, O.; Verheijen, M. A.; Bakkers, E. P. A. M. Nano Letters 2007, 7, (2), 367-371. 25. Bennett, B. R.; Soref, R. A.; del Alamo, J. A. Quantum Electronics, IEEE Journal of 1990, 26, (1), 113-122. 26. Mattias Borg, B.; Johansson, J.; Storm, K.; Deppert, K. Journal of Crystal Growth 2013, 366, 15-19.

ACS Paragon Plus Environment

18

Page 19 of 20

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Nano Letters

27. Algra, R. E.; Verheijen, M. A.; Borgström, M. T.; Feiner, L.-F.; Immink, G.; van Enckevort, W. J. P.; Vlieg, E.; Bakkers, E. P. A. M. Nature 2008, 456, (7220), 369-372. 28. Haruki, Y.; Takuya, H.; Naoteru, S.; Minoru, I. Japanese Journal of Applied Physics 2012, 51, (2R), 025601. 29. Stringfellow, G. B., Organometallic Vapor-Phase Epitaxy: Theory and Practice. Academic Press: San Diego, 1999.

ACS Paragon Plus Environment

19

Nano Letters

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

80x39mm (300 x 300 DPI)

ACS Paragon Plus Environment

Page 20 of 20