Home
Search
Collections
Journals
About
Contact us
My IOPscience
ZnO nanowires growth via reduction of ZnO powder by H2
This content has been downloaded from IOPscience. Please scroll down to see the full text. 2015 Nanotechnology 26 125602 (http://iopscience.iop.org/0957-4484/26/12/125602) View the table of contents for this issue, or go to the journal homepage for more
Download details: IP Address: 163.118.172.206 This content was downloaded on 19/03/2015 at 11:25
Please note that terms and conditions apply.
Nanotechnology Nanotechnology 26 (2015) 125602 (8pp)
doi:10.1088/0957-4484/26/12/125602
ZnO nanowires growth via reduction of ZnO powder by H2 Guy Burshtein1, Victor Lumelsky1 and Yeshayahu Lifshitz1,2 1
Faculty of Materials Science and Engineering Technion, Israel Institute of Technology, Haifa 3200003, Israel 2 Russel Berrie Nanotechnology Institute Technion, Israel Institute of Technology, Haifa 3200003, Israel E-mail: [email protected] and [email protected] Received 11 November 2014, revised 27 January 2015 Accepted for publication 2 February 2015 Published 5 March 2015 Abstract
A unique approach of ZnO nanowire growth mediated via reduction of ZnO by H2 is presented. It is less complex and more controllable than the conventional carbothermal method (reduction of ZnO by C). The chemical vapor deposition system employed allows precise control of all deposition parameters: (1) source and substrate temperatures, (2) carrier gas compositions, flow and pressure of several gases, (3) growth along a large range of distances from the source. In situ residual gas analysis allows real-time feedback of the process reactions. Controlled, stabilized, homogenous growth (characterized by scanning electron microscopy and x-ray diffraction) over relatively large areas is demonstrated. Keywords: ZnO nanowires, carbothermal, hydrogen reduction, CVD (Some figures may appear in colour only in the online journal) 1. Introduction
[5, 6]. The substrates are often covered by a metal catalyst which promotes the nucleation of the NWs [7, 8]. A variety of source materials are possible for ZnO NWs growth. Reported sources include: (1) ZnO powder which is thermally evaporated at high temperature (∼1400 °C) and low pressure [9], (2) Zn powder which is thermally evaporated and forms ZnO by introduction of oxygen to the carrier gas [10], and (3) a mixture of ZnO:C which produces Zn vapor by reduction of ZnO at the source and oxidation of Zn at the substrates (carbothermal reduction [11]). A mixture of ZnO:C is the most commonly used source material for ZnO NWs growth via TCVD due to its simplicity compared to the other approaches mentioned. The ZnO NWs growth utilizing the or carbothermal reactions (ZnOs + Cs → Znv + COg ZnOs + Cs → Znv + CO2g, s, v, g denote solid, vapor and gas respectively) is however difficult to control. This arises from the fact that the production of Zn vapor via the carbothermal reactions starts at much lower temperatures than that chosen for the growth process (typically 950 °C). Consequently ZnO NWs growth occurs not only when the source temperature reaches the selected temperature but also during the source heating and source cooling stages [12]. The growth time and growth conditions cannot thus be precisely controlled nor be stabilized leading to irreproducible NWs fabrication.
Semiconducting nanowires are 1D materials considered as building blocks of future electronic and optoelectronic devices due to the remarkable properties initiated by shrinking two of their dimensions below 100 nm [1, 2]. ZnO NWs are an important class of semiconducting NWS and are widely studied due to the unique set of properties of ZnO [3] including: a wide bandgap material (3.37 eV), UV lasing, piezoelectricity, pyroelectricity, biocompatibility and more. Implementation of semiconducting NWs in general and ZnO NWs in particular requires development of: (1) controlled fabrication processes in which the size, structure and the resulting properties are precisely determined, (2) processes offering homogenous growth of NWs over large areas as required by the industry. NWs can be grown by bottom up techniques applying wet and dry methods. Dry approaches are more abundant and controllable and involve transport of growth species by a carrier gas to the substrates [4]. The widespread thermal chemical vapor deposition (CVD) techniques utilize a source material which is heated to a temperature sufficient to produce growth species which are transported by a carrier gas to substrates held at a temperature chosen for the NWs growth 0957-4484/15/125602+08$33.00
1
© 2015 IOP Publishing Ltd Printed in the UK
G Burshtein et al
Nanotechnology 26 (2015) 125602
Figure 1. A schematic diagram of the experimental system.
The present work offers an alternative method for thermal CVD growth of ZnO NWs which overcomes the limitations of the carbothermal process: Zn vapor generation via reduction of ZnO by introduction of H2 to the carrier gas: ZnOs + H2g → Znv + H2Og (s, g, v denote solid, gas and vapor, respectively). Using this approach the system temperature profile (source temperature and substrates temperature) is stabilized under the pure carrier gas and H2 is admitted to the reaction chamber only for the required period so that the growth time and the growth conditions are precisely determined and stabilized. Such an approach was not studied before for ZnO NWs growth, to the best of our knowledge. This approach was used combining dedicated experiments in a highly controlled CVD system with various characterization techniques aiming at achieving a better control over the ZnO NWs growth process. The various growth parameters (source and substrate temperature, carrier gas composition, pressure and flow, H2 flow) were optimized to allow not only ZnO NWs growth under defined and stabilized conditions, but to enable homogeneous growth over a large area as well. Reported works on NWs thermal CVD growth in general and on ZnO NWs growth in particular probe the growth in a specific location over a very small area and the experimental conditions are not appropriately defined. It is thus very difficult to understand the growth process occurring in the entire chamber nor is it possible to compare the growth reports of different groups. The present work was thus employing a different methodology which dictated the construction of the growth system applied. It consisted of three elements: (1) the substrates were introduced along the entire cylindrical tube (both in the downstream and the upstream directions with respect to the source position) probing the homogeneity of the growth process over the distance from the source, (2) the role of the carrier gas pressure and flow was checked [13] (guided by simulations exploring the role of diffusion which carries the growth species both downstream and upstream and convection which carries them only downstream), (3) residual gas analysis (RGA) of the CVD gas environment was carried out during the entire growth process. This is why the growth was performed under different conditions that are specified in the experimental section including: (1) different flows (to explore the balance between diffusion and convection), (2) different source temperatures (affecting the Zn evaporation rate), (3) different gases (Ar + H2, Ar + O2 + H2) providing different Zn
oxidation processes at the substrates. In conducting the experiments we were guided by the substrate temperature, gas composition and flow conditions found appropriate for ZnO NWs growth from the carbothermal reaction in our previous experiments [13].
2. Experimental details The thermal CVD system consisted of a cylindrical quartz tube (inner diameter 29 mm, 1500 mm long) in a three zone horizontal furnace (900 mm long) pumped by a turbo system to a base pressure better than 10−6 mbar. The three zone furnace enabled a constant temperature (±10 °C) in the central zone and in most of the side zones. Figure 1 gives a scheme of the experimental system. The ZnO NWs growth was performed using a pure ZnO powder placed at the central heating zone (at temperatures of 800 or 950 °C). Al2O3 substrates, half of each one deposited by a 5 nm thick Au film, were used. The substrates were placed at different distances from the source in both the upstream and downstream side zones (heated to 800 °C). Two types of gas mixtures were used as carrier gas at a pressure of 30 mbar and flows of 9−45 sccm: (1) Ar + 2.5% H2 and (2) Ar + 2.5% H2 + 1% O2. The hydrogen and oxygen were introduced into the carrier gas only during the growth step to ensure that growth does not occur during the heating and/or cooling stages. Table 1 summarizes the growth conditions of the different experiments. The morphology of the NWs grown was investigated with a high resolution scanning electron microscope (HRSEM) and x-ray diffraction (XRD). A RGA system (Balzers Pfeiffer Prisma QME 200) was used during experiments to analyze the gases present in the growth process.
3. Results CVD growth of ZnO NWs from a ZnO source requires two steps: (1) reduction of the ZnO powder at the source, (2) oxidation of Zn species approaching the substrates. Reduction of ZnO by H2 is known to readily occur at 800–950 °C [13]. The issue of oxidation (which may be catalyst assisted) on the substrates is less trivial. It may occur either by the water vapor 2
G Burshtein et al
Nanotechnology 26 (2015) 125602
Table 1. Experimental conditions.
♯
Source material
1 2 3 4 5 6
ZnO ZnO ZnO ZnO ZnO ZnO
7
ZnO
Gas in Ar 2.5% H2 2.5% H2 2.5% H2 2.5% H2 2.5% H2 2.5% H2 + 1% O2 2.5% H2 + 1% O2
Flow (sccm)
Tube ID (mm)
Pressure (mbar)
Upstream (°C)
Source (°C)
Downstream (°C)
Growth time (min)
45 45 9 30 45 45
29 29 29 29 29 29
30 30 30 30 30 30
800 800 800 800 800 800
800 950 950 950 950 950
800 800 800 800 800 800
60 60 60 60 60 60
45
29
30
800
800
800
60
Figure 2. HRSEM micrographs of Al2O3 substrates arranged in different locations from the source downstream (indicated on the left side of the figures). The pressure is 30 mbar for 60 min at total flow rate of 45 sccm. The added gas is H2 2.5%. The tube diameter is 29 mm. The substrate temperature is 800 °C with different source temperatures of 800 and 950 °C.
generated at the source in the ZnO reduction process or by residual oxygen (which should not be existing in a leak tight system). Figure 2 shows that ZnO NWs can indeed grow via reduction of ZnO by H2 at a source temperature of both 800 and 950 °C using a H2 concentration of 2.5% in Ar under
30 mbar and 45 sccm and a substrate temperature of 800 °C. At 800 °C one observes ZnO NWs growth along the tube (distances of 110–230 mm from the source). The ZnO nanostructures grown using a source temperature of 950 °C include nanosheets and nanosails in addition to NWs due to 3
G Burshtein et al
Nanotechnology 26 (2015) 125602
Figure 3. HRSEM micrographs of Al2O3 substrates arranged downstream (a) and upstream (b) in different locations from the source, (indicated on the left side of the figures), with ZnO NWs grown at temperatures of 950 (source) and 800 °C (substrates) at 30 mbar for 60 min at Ar flow rates of 9, 30 and 45 sccm. H2 concentration of 2.5% is used. The tube diameter is 29 mm.
weight loss is ∼22%) and the balance between the supply of Zn species to the substrate region and the oxidation rate at this region is optimal so that large ZnO nanostructures are formed along the tube. Further increase of the flow reduces both the concentration and the residence time of the Zn and O2 species over the substrates so that the Zn species do not have sufficient time for oxidation and the density and length of the nanostructures decreases. It is interesting to follow the growth in the upstream direction. The low flow of 9 sccm enables
the higher Zn production rate at 950 °C. At 800 °C only NWs are grown but they have a large spread of diameters of ∼100–500 nm. Figure 3 shows the effect of flow on the formation of the ZnO NWs from reduction by H2. At a low flow of 9 sccm the removal rate of the reaction products at the source is slow so that the Zn formation rate saturates and there are no NWs grown (the weight loss of the source is only 6%). Only a thin layer of ZnO grains is formed. Increasing the flow to 30 sccm increases the overall formation rate of Zn (the 4
G Burshtein et al
Nanotechnology 26 (2015) 125602
Figure 3. (Continued.)
diffusion of Zn in the upstream direction so that some ZnO nanostructures are formed. At increased flows convection is much more dominant than diffusion, Zn species are not reaching upstream positions and no upstream growth occurs. So far the experiments probed the growth of ZnO NWs by reduction via H2 only relying on the produced water vapor or residual O2 for oxidation of the Zn on the substrates. Figure 4(a) shows the effect of introduction of 1% O2 to the Ar + 2.5% H2 carrier gas. The addition of O2 definitely increased the oxidation rate at the substrates causing massive formation of ZnO nanostructures (NWs, nanosheets and nanosails) along the entire length of the reaction tube (380 mm from the source). XRD analysis of the products (figure 4(b)) confirms the formation of wurtzite ZnO. For most applications the formation of a host of different nanostructures is undesired and a homogeneous growth of NWs is requested. This can be achieved by reducing the source temperature to 800 °C and thus reducing the reaction rate and the total weight loss of the source material to ∼3%. Figure 5 shows that homogeneous growth of ZnO NWs can be achieved under these conditions (with narrower (∼50–100 nm in diameter), longer, and more uniform NWs along their axis)
up to a distance of 200 mm from the source. Above 200 mm the density and length of the NWs decrease. This is likely due to decrease of the Zn concentration with distance (e.g. caused by sedimentation of Zn). Further optimization of the growth uniformity over larger distances requires additional tuning by the process conditions applied. The RGA data is helpful to follow the reactions occurring during the NWs growth. Figure 6 shows the RGA data related to the experiment of figure 5 (Ar + 2.5% H2 + 1% O2, 45 sccm, source temperature 800 °C). The introduction of H2 and O2 at the beginning of the process is associated with the increase of both mass 2 and mass 32. The reduction of ZnO by H2 leads to the formation of water and consumption of H2. Upon the end of the process the H2, and O2 supply is stopped and the formation of water vapor ceases as well.
4. Discussion CVD growth of NWs involves generation of growth species in the source region (source reactions), their transport to the substrate region, and finally nucleation and growth of NWs in 5
G Burshtein et al
Nanotechnology 26 (2015) 125602
Figure 4. (a) HRSEM micrographs of Al2O3 substrates arranged in different locations from the source downstream (indicated on the left side of the figures), with ZnO NWs grown at temperatures of 950 (source) and 800 °C (substrates) at 30 mbar for 60 min at total flow rate of 45 sccm. The added gas is H2 2.5% + O2 1%. Tube diameter of 29 mm was used. (b) XRD of substrate at 110 mm position.
conditions lead to downstream growth and fast establishment of steady state conditions as far as the growth species distribution is concerned. Optimization of ZnO NWs growth conditions thus involves control of: (1) the growth species (typically Zn) generation rate in the source (through the source temperature), (2) the growth species distribution balancing between diffusion and convection (mainly through tuning of the flow rate), (3) the oxidation of Zn at the target region and nucleation and growth of NWs (via supply of oxidizing species, provision of nucleation sites and tuning the substrate temperature). The carbothermal reaction has the advantage of being able to produce Zn at medium temperatures (say 950 °C). It is
the substrate region (substrate reactions). In the case of ZnO NWs growth both the carbothermal and the hydrogen processes involve reduction of ZnO, forming Zn species in the source region and oxidation of Zn on the nucleation sites of the substrate (that may be catalyst assisted). The transport of growth species is dictated by the equilibrium between diffusion (carrying the growth species to both downstream and upstream directions with respect to the source) and convection (only downstream transport is possible). At low flow conditions diffusion is dominant so that the growth species are transported in the entire system, both downstream and upstream growth are possible (see figure 3) and steady state conditions develop after a long time. In contrast, high flow 6
G Burshtein et al
Nanotechnology 26 (2015) 125602
Figure 5. HRSEM micrographs of Al2O3 substrates arranged in different locations from the source downstream (indicated on the left side of the figures), with ZnO NWs grown at source and substrate temperatures of 800 °C. The pressure is 30 mbar for 60 min at total flow rate of 45 sccm. The added gas is H2 2.5% + O2 1%. A 29 mm ID tube was used. For each position a low magnification micrograph is also presented (right). 7
G Burshtein et al
Nanotechnology 26 (2015) 125602
5. Conclusions In summary, the growth of ZnO NWs using H2 reduction in thermal CVD systems was assessed. Growth experiments at different conditions (source temperature, flow rate and external supply of oxygen as an oxidizing agent) were followed by characterizations using HRSEM, XRD and RGA. It was shown that reduction by H2 of a ZnO powder can be used for ZnO NWs growth without an additional oxidizing element. Introduction of external O2 can however enhance the oxidation of the Zn growth species on the substrates and lead to homogeneous growth on large area by optimization of the growth conditions. Reduction of ZnO by H2 provides a simple and highly controlled alternative to the commonly used carbothermal reduction. The approach suggested herein is helpful for the development of controllable, scalable ZnO NWs growth processes suitable for industrial applications. Figure 6. RGA data for the experiment of ZnO NWs growth from
hydrogen reduction of ZnO using hydrogen reduction (the growth conditions are indicated above). The NWs grown in this experiment are shown in figure 5.
Acknowledgments This work was partially supported by the German Israeli Cooperation program (DIP) project K 6.1 and by the Russel Berrie Nanotechnology Institute.
however complex and has many drawbacks for establishing controllable and reproducible processes. One drawback is that the reduction of ZnO by C is not an on–off process as far as the source temperature is concerned. Zn species are generated over a large temperature range. Zn is thus generated during both the heating and cooling periods of the source, steady state growth conditions cannot be achieved and the growth period cannot be defined and controlled. The use of the ‘pressure shutter’ [14] (introduction of high pressure to suppress growth and supply of the appropriate pressure only at the growth period) or ‘reverse flow’ [13] (providing the carrier gas flow in the upstream direction first and switching it to the downstream flow only at the growth period) are insufficient to guarantee growth at a defined period under stable conditions due to diffusion effects. Another drawback of the carbothermal reaction is that it is impossible to isolate between the source reactions and the substrate reactions. The carbothermal reduction produces oxidizing products (CO and CO2) which are carried to the substrate region. On the other hand addition of O2 or CO2 to the carrier gas in order to enhance oxidation in the substrate region affects the reduction reactions in the source region. The reduction of ZnO by H2 is a much simpler process which allows for a much better independent control of the reduction reactions in the source region and the oxidation reactions in the target region. Not less important is the possibility to introduce H2 at a precise moment and thus define the growth duration and achieve steady growth conditions. Reduction of ZnO by H2 thus provides a simple, controllable method for reproducible ZnO growth over large areas as demonstrated in the present work. This method is definitely scalable for industrial applications.
References [1] Barth S 2010 Prog. Mater. Sci. 55 563 [2] Kuchibhatla S V N T, Karakoti A S, Bera D and Seal S 2007 Prog. Mater. Sci. 52 699 [3] Jagadish C and Pearton S 2006 Zinc Oxide Bulk, Thin Films and Nanostructures (Amsterdam: Elsevier) pp 1–20 [4] Fan H J, Werner P and Zacharias M 2006 Small 2 700 [5] Wan H and Ruda H E 2010 J. Mater. Sci., Mater. Electron. 21 1014 [6] Amarilio-Burshtein I, Tamir S and Lifshitz Y 2010 Appl. Phys. Lett. 96 103104-1 [7] Zhang Z, Wang S J, Yu T and Wu T 2007 J. Phys. Chem. C 111 17500 [8] Guo D L, Huang X, Xing G Z, Zhang Z, Li G P, He M, Zhang H, Chen H and Wu T 2011 Phys. Rev. B 83 045403 [9] Pan Z W, Dai Z R and Wang Z L 2001 Science 291 1947 [10] Yan H, He R, Pham J and Yang P 2003 Adv. Mater. 15 402 [11] Yang P, Yan H, Mao S, Russo R, Johnson J, Saykally R, Morris N, Pham J, He R and Choi H J 2002 Adv. Funct. Mater. 12 323 [12] Wongchoosuk C, Subannajui K, Menzel A, Amarilio-Burshtein I, Tamir S, Lifshitz S and Zacharias M 2011 J. Phys. Chem. C 115 757 [13] Menzel A, Goldberg R, Burshtein G, Lumelsky V, Subannajui K, Zacharias M and Lifshitz Y 2012 J. Phys. Chem. C 116 5524 [14] Dalal S H, Baptista D L, Teo K B K, Lacerda R G, Jefferson D A and Milne W I 2006 Nanotechnology 17 4811
8
Search
Collections
Journals
About
Contact us
My IOPscience
ZnO nanowires growth via reduction of ZnO powder by H2
This content has been downloaded from IOPscience. Please scroll down to see the full text. 2015 Nanotechnology 26 125602 (http://iopscience.iop.org/0957-4484/26/12/125602) View the table of contents for this issue, or go to the journal homepage for more
Download details: IP Address: 163.118.172.206 This content was downloaded on 19/03/2015 at 11:25
Please note that terms and conditions apply.
Nanotechnology Nanotechnology 26 (2015) 125602 (8pp)
doi:10.1088/0957-4484/26/12/125602
ZnO nanowires growth via reduction of ZnO powder by H2 Guy Burshtein1, Victor Lumelsky1 and Yeshayahu Lifshitz1,2 1
Faculty of Materials Science and Engineering Technion, Israel Institute of Technology, Haifa 3200003, Israel 2 Russel Berrie Nanotechnology Institute Technion, Israel Institute of Technology, Haifa 3200003, Israel E-mail: [email protected] and [email protected] Received 11 November 2014, revised 27 January 2015 Accepted for publication 2 February 2015 Published 5 March 2015 Abstract
A unique approach of ZnO nanowire growth mediated via reduction of ZnO by H2 is presented. It is less complex and more controllable than the conventional carbothermal method (reduction of ZnO by C). The chemical vapor deposition system employed allows precise control of all deposition parameters: (1) source and substrate temperatures, (2) carrier gas compositions, flow and pressure of several gases, (3) growth along a large range of distances from the source. In situ residual gas analysis allows real-time feedback of the process reactions. Controlled, stabilized, homogenous growth (characterized by scanning electron microscopy and x-ray diffraction) over relatively large areas is demonstrated. Keywords: ZnO nanowires, carbothermal, hydrogen reduction, CVD (Some figures may appear in colour only in the online journal) 1. Introduction
[5, 6]. The substrates are often covered by a metal catalyst which promotes the nucleation of the NWs [7, 8]. A variety of source materials are possible for ZnO NWs growth. Reported sources include: (1) ZnO powder which is thermally evaporated at high temperature (∼1400 °C) and low pressure [9], (2) Zn powder which is thermally evaporated and forms ZnO by introduction of oxygen to the carrier gas [10], and (3) a mixture of ZnO:C which produces Zn vapor by reduction of ZnO at the source and oxidation of Zn at the substrates (carbothermal reduction [11]). A mixture of ZnO:C is the most commonly used source material for ZnO NWs growth via TCVD due to its simplicity compared to the other approaches mentioned. The ZnO NWs growth utilizing the or carbothermal reactions (ZnOs + Cs → Znv + COg ZnOs + Cs → Znv + CO2g, s, v, g denote solid, vapor and gas respectively) is however difficult to control. This arises from the fact that the production of Zn vapor via the carbothermal reactions starts at much lower temperatures than that chosen for the growth process (typically 950 °C). Consequently ZnO NWs growth occurs not only when the source temperature reaches the selected temperature but also during the source heating and source cooling stages [12]. The growth time and growth conditions cannot thus be precisely controlled nor be stabilized leading to irreproducible NWs fabrication.
Semiconducting nanowires are 1D materials considered as building blocks of future electronic and optoelectronic devices due to the remarkable properties initiated by shrinking two of their dimensions below 100 nm [1, 2]. ZnO NWs are an important class of semiconducting NWS and are widely studied due to the unique set of properties of ZnO [3] including: a wide bandgap material (3.37 eV), UV lasing, piezoelectricity, pyroelectricity, biocompatibility and more. Implementation of semiconducting NWs in general and ZnO NWs in particular requires development of: (1) controlled fabrication processes in which the size, structure and the resulting properties are precisely determined, (2) processes offering homogenous growth of NWs over large areas as required by the industry. NWs can be grown by bottom up techniques applying wet and dry methods. Dry approaches are more abundant and controllable and involve transport of growth species by a carrier gas to the substrates [4]. The widespread thermal chemical vapor deposition (CVD) techniques utilize a source material which is heated to a temperature sufficient to produce growth species which are transported by a carrier gas to substrates held at a temperature chosen for the NWs growth 0957-4484/15/125602+08$33.00
1
© 2015 IOP Publishing Ltd Printed in the UK
G Burshtein et al
Nanotechnology 26 (2015) 125602
Figure 1. A schematic diagram of the experimental system.
The present work offers an alternative method for thermal CVD growth of ZnO NWs which overcomes the limitations of the carbothermal process: Zn vapor generation via reduction of ZnO by introduction of H2 to the carrier gas: ZnOs + H2g → Znv + H2Og (s, g, v denote solid, gas and vapor, respectively). Using this approach the system temperature profile (source temperature and substrates temperature) is stabilized under the pure carrier gas and H2 is admitted to the reaction chamber only for the required period so that the growth time and the growth conditions are precisely determined and stabilized. Such an approach was not studied before for ZnO NWs growth, to the best of our knowledge. This approach was used combining dedicated experiments in a highly controlled CVD system with various characterization techniques aiming at achieving a better control over the ZnO NWs growth process. The various growth parameters (source and substrate temperature, carrier gas composition, pressure and flow, H2 flow) were optimized to allow not only ZnO NWs growth under defined and stabilized conditions, but to enable homogeneous growth over a large area as well. Reported works on NWs thermal CVD growth in general and on ZnO NWs growth in particular probe the growth in a specific location over a very small area and the experimental conditions are not appropriately defined. It is thus very difficult to understand the growth process occurring in the entire chamber nor is it possible to compare the growth reports of different groups. The present work was thus employing a different methodology which dictated the construction of the growth system applied. It consisted of three elements: (1) the substrates were introduced along the entire cylindrical tube (both in the downstream and the upstream directions with respect to the source position) probing the homogeneity of the growth process over the distance from the source, (2) the role of the carrier gas pressure and flow was checked [13] (guided by simulations exploring the role of diffusion which carries the growth species both downstream and upstream and convection which carries them only downstream), (3) residual gas analysis (RGA) of the CVD gas environment was carried out during the entire growth process. This is why the growth was performed under different conditions that are specified in the experimental section including: (1) different flows (to explore the balance between diffusion and convection), (2) different source temperatures (affecting the Zn evaporation rate), (3) different gases (Ar + H2, Ar + O2 + H2) providing different Zn
oxidation processes at the substrates. In conducting the experiments we were guided by the substrate temperature, gas composition and flow conditions found appropriate for ZnO NWs growth from the carbothermal reaction in our previous experiments [13].
2. Experimental details The thermal CVD system consisted of a cylindrical quartz tube (inner diameter 29 mm, 1500 mm long) in a three zone horizontal furnace (900 mm long) pumped by a turbo system to a base pressure better than 10−6 mbar. The three zone furnace enabled a constant temperature (±10 °C) in the central zone and in most of the side zones. Figure 1 gives a scheme of the experimental system. The ZnO NWs growth was performed using a pure ZnO powder placed at the central heating zone (at temperatures of 800 or 950 °C). Al2O3 substrates, half of each one deposited by a 5 nm thick Au film, were used. The substrates were placed at different distances from the source in both the upstream and downstream side zones (heated to 800 °C). Two types of gas mixtures were used as carrier gas at a pressure of 30 mbar and flows of 9−45 sccm: (1) Ar + 2.5% H2 and (2) Ar + 2.5% H2 + 1% O2. The hydrogen and oxygen were introduced into the carrier gas only during the growth step to ensure that growth does not occur during the heating and/or cooling stages. Table 1 summarizes the growth conditions of the different experiments. The morphology of the NWs grown was investigated with a high resolution scanning electron microscope (HRSEM) and x-ray diffraction (XRD). A RGA system (Balzers Pfeiffer Prisma QME 200) was used during experiments to analyze the gases present in the growth process.
3. Results CVD growth of ZnO NWs from a ZnO source requires two steps: (1) reduction of the ZnO powder at the source, (2) oxidation of Zn species approaching the substrates. Reduction of ZnO by H2 is known to readily occur at 800–950 °C [13]. The issue of oxidation (which may be catalyst assisted) on the substrates is less trivial. It may occur either by the water vapor 2
G Burshtein et al
Nanotechnology 26 (2015) 125602
Table 1. Experimental conditions.
♯
Source material
1 2 3 4 5 6
ZnO ZnO ZnO ZnO ZnO ZnO
7
ZnO
Gas in Ar 2.5% H2 2.5% H2 2.5% H2 2.5% H2 2.5% H2 2.5% H2 + 1% O2 2.5% H2 + 1% O2
Flow (sccm)
Tube ID (mm)
Pressure (mbar)
Upstream (°C)
Source (°C)
Downstream (°C)
Growth time (min)
45 45 9 30 45 45
29 29 29 29 29 29
30 30 30 30 30 30
800 800 800 800 800 800
800 950 950 950 950 950
800 800 800 800 800 800
60 60 60 60 60 60
45
29
30
800
800
800
60
Figure 2. HRSEM micrographs of Al2O3 substrates arranged in different locations from the source downstream (indicated on the left side of the figures). The pressure is 30 mbar for 60 min at total flow rate of 45 sccm. The added gas is H2 2.5%. The tube diameter is 29 mm. The substrate temperature is 800 °C with different source temperatures of 800 and 950 °C.
generated at the source in the ZnO reduction process or by residual oxygen (which should not be existing in a leak tight system). Figure 2 shows that ZnO NWs can indeed grow via reduction of ZnO by H2 at a source temperature of both 800 and 950 °C using a H2 concentration of 2.5% in Ar under
30 mbar and 45 sccm and a substrate temperature of 800 °C. At 800 °C one observes ZnO NWs growth along the tube (distances of 110–230 mm from the source). The ZnO nanostructures grown using a source temperature of 950 °C include nanosheets and nanosails in addition to NWs due to 3
G Burshtein et al
Nanotechnology 26 (2015) 125602
Figure 3. HRSEM micrographs of Al2O3 substrates arranged downstream (a) and upstream (b) in different locations from the source, (indicated on the left side of the figures), with ZnO NWs grown at temperatures of 950 (source) and 800 °C (substrates) at 30 mbar for 60 min at Ar flow rates of 9, 30 and 45 sccm. H2 concentration of 2.5% is used. The tube diameter is 29 mm.
weight loss is ∼22%) and the balance between the supply of Zn species to the substrate region and the oxidation rate at this region is optimal so that large ZnO nanostructures are formed along the tube. Further increase of the flow reduces both the concentration and the residence time of the Zn and O2 species over the substrates so that the Zn species do not have sufficient time for oxidation and the density and length of the nanostructures decreases. It is interesting to follow the growth in the upstream direction. The low flow of 9 sccm enables
the higher Zn production rate at 950 °C. At 800 °C only NWs are grown but they have a large spread of diameters of ∼100–500 nm. Figure 3 shows the effect of flow on the formation of the ZnO NWs from reduction by H2. At a low flow of 9 sccm the removal rate of the reaction products at the source is slow so that the Zn formation rate saturates and there are no NWs grown (the weight loss of the source is only 6%). Only a thin layer of ZnO grains is formed. Increasing the flow to 30 sccm increases the overall formation rate of Zn (the 4
G Burshtein et al
Nanotechnology 26 (2015) 125602
Figure 3. (Continued.)
diffusion of Zn in the upstream direction so that some ZnO nanostructures are formed. At increased flows convection is much more dominant than diffusion, Zn species are not reaching upstream positions and no upstream growth occurs. So far the experiments probed the growth of ZnO NWs by reduction via H2 only relying on the produced water vapor or residual O2 for oxidation of the Zn on the substrates. Figure 4(a) shows the effect of introduction of 1% O2 to the Ar + 2.5% H2 carrier gas. The addition of O2 definitely increased the oxidation rate at the substrates causing massive formation of ZnO nanostructures (NWs, nanosheets and nanosails) along the entire length of the reaction tube (380 mm from the source). XRD analysis of the products (figure 4(b)) confirms the formation of wurtzite ZnO. For most applications the formation of a host of different nanostructures is undesired and a homogeneous growth of NWs is requested. This can be achieved by reducing the source temperature to 800 °C and thus reducing the reaction rate and the total weight loss of the source material to ∼3%. Figure 5 shows that homogeneous growth of ZnO NWs can be achieved under these conditions (with narrower (∼50–100 nm in diameter), longer, and more uniform NWs along their axis)
up to a distance of 200 mm from the source. Above 200 mm the density and length of the NWs decrease. This is likely due to decrease of the Zn concentration with distance (e.g. caused by sedimentation of Zn). Further optimization of the growth uniformity over larger distances requires additional tuning by the process conditions applied. The RGA data is helpful to follow the reactions occurring during the NWs growth. Figure 6 shows the RGA data related to the experiment of figure 5 (Ar + 2.5% H2 + 1% O2, 45 sccm, source temperature 800 °C). The introduction of H2 and O2 at the beginning of the process is associated with the increase of both mass 2 and mass 32. The reduction of ZnO by H2 leads to the formation of water and consumption of H2. Upon the end of the process the H2, and O2 supply is stopped and the formation of water vapor ceases as well.
4. Discussion CVD growth of NWs involves generation of growth species in the source region (source reactions), their transport to the substrate region, and finally nucleation and growth of NWs in 5
G Burshtein et al
Nanotechnology 26 (2015) 125602
Figure 4. (a) HRSEM micrographs of Al2O3 substrates arranged in different locations from the source downstream (indicated on the left side of the figures), with ZnO NWs grown at temperatures of 950 (source) and 800 °C (substrates) at 30 mbar for 60 min at total flow rate of 45 sccm. The added gas is H2 2.5% + O2 1%. Tube diameter of 29 mm was used. (b) XRD of substrate at 110 mm position.
conditions lead to downstream growth and fast establishment of steady state conditions as far as the growth species distribution is concerned. Optimization of ZnO NWs growth conditions thus involves control of: (1) the growth species (typically Zn) generation rate in the source (through the source temperature), (2) the growth species distribution balancing between diffusion and convection (mainly through tuning of the flow rate), (3) the oxidation of Zn at the target region and nucleation and growth of NWs (via supply of oxidizing species, provision of nucleation sites and tuning the substrate temperature). The carbothermal reaction has the advantage of being able to produce Zn at medium temperatures (say 950 °C). It is
the substrate region (substrate reactions). In the case of ZnO NWs growth both the carbothermal and the hydrogen processes involve reduction of ZnO, forming Zn species in the source region and oxidation of Zn on the nucleation sites of the substrate (that may be catalyst assisted). The transport of growth species is dictated by the equilibrium between diffusion (carrying the growth species to both downstream and upstream directions with respect to the source) and convection (only downstream transport is possible). At low flow conditions diffusion is dominant so that the growth species are transported in the entire system, both downstream and upstream growth are possible (see figure 3) and steady state conditions develop after a long time. In contrast, high flow 6
G Burshtein et al
Nanotechnology 26 (2015) 125602
Figure 5. HRSEM micrographs of Al2O3 substrates arranged in different locations from the source downstream (indicated on the left side of the figures), with ZnO NWs grown at source and substrate temperatures of 800 °C. The pressure is 30 mbar for 60 min at total flow rate of 45 sccm. The added gas is H2 2.5% + O2 1%. A 29 mm ID tube was used. For each position a low magnification micrograph is also presented (right). 7
G Burshtein et al
Nanotechnology 26 (2015) 125602
5. Conclusions In summary, the growth of ZnO NWs using H2 reduction in thermal CVD systems was assessed. Growth experiments at different conditions (source temperature, flow rate and external supply of oxygen as an oxidizing agent) were followed by characterizations using HRSEM, XRD and RGA. It was shown that reduction by H2 of a ZnO powder can be used for ZnO NWs growth without an additional oxidizing element. Introduction of external O2 can however enhance the oxidation of the Zn growth species on the substrates and lead to homogeneous growth on large area by optimization of the growth conditions. Reduction of ZnO by H2 provides a simple and highly controlled alternative to the commonly used carbothermal reduction. The approach suggested herein is helpful for the development of controllable, scalable ZnO NWs growth processes suitable for industrial applications. Figure 6. RGA data for the experiment of ZnO NWs growth from
hydrogen reduction of ZnO using hydrogen reduction (the growth conditions are indicated above). The NWs grown in this experiment are shown in figure 5.
Acknowledgments This work was partially supported by the German Israeli Cooperation program (DIP) project K 6.1 and by the Russel Berrie Nanotechnology Institute.
however complex and has many drawbacks for establishing controllable and reproducible processes. One drawback is that the reduction of ZnO by C is not an on–off process as far as the source temperature is concerned. Zn species are generated over a large temperature range. Zn is thus generated during both the heating and cooling periods of the source, steady state growth conditions cannot be achieved and the growth period cannot be defined and controlled. The use of the ‘pressure shutter’ [14] (introduction of high pressure to suppress growth and supply of the appropriate pressure only at the growth period) or ‘reverse flow’ [13] (providing the carrier gas flow in the upstream direction first and switching it to the downstream flow only at the growth period) are insufficient to guarantee growth at a defined period under stable conditions due to diffusion effects. Another drawback of the carbothermal reaction is that it is impossible to isolate between the source reactions and the substrate reactions. The carbothermal reduction produces oxidizing products (CO and CO2) which are carried to the substrate region. On the other hand addition of O2 or CO2 to the carrier gas in order to enhance oxidation in the substrate region affects the reduction reactions in the source region. The reduction of ZnO by H2 is a much simpler process which allows for a much better independent control of the reduction reactions in the source region and the oxidation reactions in the target region. Not less important is the possibility to introduce H2 at a precise moment and thus define the growth duration and achieve steady growth conditions. Reduction of ZnO by H2 thus provides a simple, controllable method for reproducible ZnO growth over large areas as demonstrated in the present work. This method is definitely scalable for industrial applications.
References [1] Barth S 2010 Prog. Mater. Sci. 55 563 [2] Kuchibhatla S V N T, Karakoti A S, Bera D and Seal S 2007 Prog. Mater. Sci. 52 699 [3] Jagadish C and Pearton S 2006 Zinc Oxide Bulk, Thin Films and Nanostructures (Amsterdam: Elsevier) pp 1–20 [4] Fan H J, Werner P and Zacharias M 2006 Small 2 700 [5] Wan H and Ruda H E 2010 J. Mater. Sci., Mater. Electron. 21 1014 [6] Amarilio-Burshtein I, Tamir S and Lifshitz Y 2010 Appl. Phys. Lett. 96 103104-1 [7] Zhang Z, Wang S J, Yu T and Wu T 2007 J. Phys. Chem. C 111 17500 [8] Guo D L, Huang X, Xing G Z, Zhang Z, Li G P, He M, Zhang H, Chen H and Wu T 2011 Phys. Rev. B 83 045403 [9] Pan Z W, Dai Z R and Wang Z L 2001 Science 291 1947 [10] Yan H, He R, Pham J and Yang P 2003 Adv. Mater. 15 402 [11] Yang P, Yan H, Mao S, Russo R, Johnson J, Saykally R, Morris N, Pham J, He R and Choi H J 2002 Adv. Funct. Mater. 12 323 [12] Wongchoosuk C, Subannajui K, Menzel A, Amarilio-Burshtein I, Tamir S, Lifshitz S and Zacharias M 2011 J. Phys. Chem. C 115 757 [13] Menzel A, Goldberg R, Burshtein G, Lumelsky V, Subannajui K, Zacharias M and Lifshitz Y 2012 J. Phys. Chem. C 116 5524 [14] Dalal S H, Baptista D L, Teo K B K, Lacerda R G, Jefferson D A and Milne W I 2006 Nanotechnology 17 4811
8
