Study of nanostructure growth with nanoscale apex induced by femtosecond laser irradiation at megahertz repetition rate
© Patel et al.; licensee Springer. 2013
Received: 7 February 2013
Accepted: 6 April 2013
Published: 22 April 2013
Leaf-like nanostructures with nanoscale apex are induced on dielectric target surfaces by high-repetition-rate femtosecond laser irradiation in ambient conditions. We have recently developed this unique technique to grow leaf-like nanostructures with such interesting geometry without the use of any catalyst. It was found to be possible only in the presence of background nitrogen gas flow. In this synthesis method, the target serves as the source for building material as well as the substrate upon which these nanostructures can grow. In our investigation, it was found that there are three possible kinds of nanotips that can grow on target surfaces. In this report, we have presented the study of the growth mechanisms of such leaf-like nanostructures under various conditions such as different laser pulse widths, pulse repetition rates, dwell times, and laser polarizations. We observed a clear transformation in the kind of nanotips that grew for the given laser conditions.
Nanostructures with nanoscale apex have become the center of attraction for many researchers around the world. These nanostructures have been widely named as nanotips, nanocones, nanonails, nanopencils, nanojets, and nanoneedles. They are considered to be one-dimensional nanostructures with a significantly large surface-to-volume ratio which is very desirable for the development of various novel devices. These nanostructures provide unique optical, electronic, mechanical, chemical, and other properties that can be very useful for the improvement of electronics interconnects, scanning probes, nanoelectronics, photovoltaic devices, electron field microemitters, light-emitting diodes, and photo-detectors [1, 2]. Until now, such nanostructures have been mainly generated from materials such as ZnO, AlN, single and polycrystalline silicon, gold, and carbon whose growth is dependent on the crystallographic orientation. These nanostructures have been synthesized by techniques such as thermal evaporation, various types of chemical vapor deposition, resonance plasma etching, and chemical etching [2–8]. The aforementioned techniques require a long processing time, multiple steps, catalyst-assisted growth, high processing temperatures, very sophisticated equipment, vacuum, and clean room operations.
In the past few years, various types of lasers have also been utilized to produce micronanostructures with sharp ends (nanobumps, nanojets, nanoprotrusions) from the irradiation of thin metal films and bulk materials using tightly focused laser beams. Such sharp nanojet structures have been produced on gold thin films by irradiation of single nano- or femtosecond laser pulse in ambient or under low-vacuum conditions using circular laser spots . In most of these cases, the gold films with certain thicknesses were deposited onto borosilicate glass or single-crystal silicon substrates by RF sputtering with the help of in situ coating of adhesion layers [9, 10]. In these techniques, for each laser pulse interaction with the film, only one nanostructure is produced at a time, and the distance between two laser incident spots on the film has to be maintained at a certain value to avoid potential rupture of the film and the damage of the previously formed nanostructure via intersection of laser irradiation spots . This eventually limits the number of nanostructures that can be produced on a surface area of the target. The study of these nanostructures for various parameters has been conducted by various researchers on various metal films [9–12]. The number of laser pulses that can be applied onto a particular spot on the target film is limited due to the fact that multiple laser pulses could ablate all the film material from the irradiation spot and could eventually start ablating the substrate surface. However, the multiple laser pulses have been used to produce sharp spikes on bulk silicon surfaces in vacuum chamber filled with 500 Torr of Cl2, SF6, N2, or He gas . They have reported that the silicon surface irradiated in SF6 and Cl2 gas background exhibits the growth of sharp spikes roughly aligned in rows whereas in the case of vacuum, N2, or He gas background, very blunt spikes with irregular sides and rounded tops with much larger tip diameter are formed. The chemical reaction between silicon surface and the surrounding gas has been suggested to be the driving factor for the formation of this sharp spikes, although the exact formation mechanism is still unknown . In another investigation, the silicon spikes have also been produced by femtosecond laser irradiation in submerged condition in water . The spikes produced in this method are one to two orders of magnitude smaller than spikes induced in . The silicon wafer is placed in a glass container filled with distilled water which is mounted on a three-axis translation stage. In their investigation, they found that for each incident laser pulse onto the silicon surface, two to three microbubbles are created in the water corresponding to which the same number of ripple-like structures are created onto the silicon surface. As more laser pulses are applied, more numbers of ripple structures are created which start to overlap with each other and roughens the silicon surface. These interactions result in generation of many submicrometer bead-like structures on silicon surface which eventually sharpen and grow into spikes through preferential removal of material around the beads by laser-assisted etching.
In this paper, we report the study of how the formation of leaf-like nanotips on femtosecond laser-irradiated transparent dielectric target is influenced due to various femtosecond laser pulse widths, repetition rates, machining dwell time, and laser polarization. Understanding the changes in generated nanotips will help us pick the right combinations of laser parameters to grow the desired amount and kind of nanotips over the large surface area of dielectric targets.
The experiments were performed on plain microscopic slide glass with composition of 60% to 75 wt.% SiO2, 5% to 12 wt.% CaO, and 12% to 18 wt.% Na2O. A direct-diode-pumped Yb-doped fiber amplifier/oscillator system (wavelength, λ = 1,030 nm) capable of delivering a maximum average output of 16 W was used as a femtosecond laser source to irradiate targets with thickness of 0.90 to 1.0 mm. The laser intensity profile beam was focused into a spot (full width at half-maximum) diameter of 10 μm on the target surface using a telecentric lens of 100-mm effective focal length. The same setup was used to perform these experiments as reported in a previous paper done by our research group . However, for these experiments, a square bracket was placed in front of the target surface which holds six nozzles providing continuous flow of nitrogen gas. The machining was performed in the form of 26 × 26 arrays of microholes for various femtosecond laser parameters. We investigated the effect of three different pulse widths (214, 428, and 714 fs) on the generation of nanotips for a repetition rate of 13 MHz at a dwell time of 0.5 ms. The effect of various laser pulse repetition rates (4, 8, and 13 MHz) and different dwell times was also investigated on glass samples. All the aforementioned experiments were done by circular polarization of laser pulses. We also examined how different (linear, p-) polarizations would change the growth of nanotips on the target surface. The linear (p-) polarization of the beam was achieved by placing a half-wave plate in front of the focusing lens. The laser-irradiated glass samples have been analyzed by SEM.
Results and discussion
Summary of effects of laser conditions to tip growth
Effects on nanotip growth
Short pulses yield narrow long tips
Higher repetition rate promotes the growth of dense, oriented narrow nanotips
Longer dwell time increases the population of nanotips. However, beyond an optimum dwell time, over heating will remelt the newly formed nanotips
Linear (p-) polarization increases the population of nanotips
Effect of pulse width
There are two mechanisms responsible for laser-induced optical breakdown of materials: multiphoton absorption and avalanche ionization. Multiphoton absorption results when a molecule absorbs the required amount of photons simultaneously to get ionized, which has proven to be the main mechanism for breakdown in the low-femtosecond regime . In our experiments, the investigated pulse widths fall above the low-femtosecond regime where the combination of both mechanisms is believed to be responsible for the breakdown. Multiphoton ionization is responsible for the initial generation of electrons which are further heated by incoming portion of the pulse resulting in avalanche ionization and rapid plasma formation . The initial part of the pulse produces free-electron plasma which can absorb the later part more efficiently and/or behave as a mirror and reflect most of the incident energy [17, 19, 20]. Every material has its unique optical damage fluence, but all the pure dielectrics demonstrate similar behavior in all ranges of pulse width as observed for SiO2. Stuart et al. investigated the threshold fluence for fused silica and CaF2 with laser pulses in the range 270 fs ≤ τ ≤ 1 ns . They discovered that the damage threshold decreased with the decrease of the pulse width.
Even though we did not work in the picosecond pulse duration regime, we obtained similar result as we increased the pulse width in the femtosecond regime. Figure 3 shows the SEM images of the microholes drilled by femtosecond laser pulses at 13-MHz repetition rate for 0.5-ms dwell time with pulse widths of 214 and 714 fs, respectively. The diameters of these microholes are approximately on average 12 and 21 μm, respectively. The size of microhole represent the amount of material removed from the target; larger diameter means larger amount of material removal compare to smaller hole diameter. The life span of the plasma is also an important factor. In the current investigation, the turbulence created in the plasma due to the interactions between nitrogen gas and plasma species lengthens the plasma life. Since the longer pulses spend a significant portion of their duration traveling through previously formed plasma, as depicted in Figure 2, the energy transmitted via longer pulse is not enough to ablate the material upon contact with the target material. Rather, this transmitted energy gets stored in the top part of the lattice and gets transferred into the bulk in all directions, making the target temperature rise in the area surrounding the irradiated spot. This makes molecules to become loose to form a larger pool of molten material. As a result, the subsequent longer pulses expel large particles and droplets into the plasma upon contacting the molten pool. On the contrary, the interaction of the short pulses with the target surface does not rise as much high temperature which creates shallow molten pool. Hence, the material removed from the target is composed of smaller particles and droplets. The size of the plasma species and the temperature rise of the target surface greatly affect the type of nanotips that grow on the target surface.
In our study, the nitrogen gas flow generates extra turbulence in expanding the plasma. As a result, the plasma species experience many collisions with each other, resulting in the formation of larger droplets. The longer pulse creates high temperature in the target surface, resulting in most of the redeposited droplets being spread into the film before getting cooled down into their original shape using nitrogen gas. There are still chances of forming smaller droplets in the plasma vaporization since plasma species interaction is very random. However, the smaller droplets are most likely to get dissolved into the surface molten layer because of the higher target surface and molten film temperatures. At 428-fs pulse width, as seen in Figure 5b, there are a significant number of nanotips growing from the molten film. When the laser pulse width was further increased to 714 fs, a very small number of nanotips are found to be growing even though it formed from the molten target material, as observed in Figure 5c. This might be due to the fact that during the 714-fs pulse interaction with the target surface, a very large amount of molten material is created which gets ejected into the plasma as well as pushed around the drilled hole due to the shock waves in the plasma. As a result, very short nanotips are observed to be growing from relatively large liquid volume of molten glass, as seen in Figure 5c.
Effect of laser pulse repetition rate
We have studied three different pulse repetition rates (13, 8, and 4 MHz) in our experiments. As the repetition rate is reduced, the time separation between two pulses increases which eventually changes how the pulse energy is being transferred into the target and being used to ablate the material from the target. If the time gap between two pulses is less than the time required for heat to diffuse out of the focal volume for a typical glass, then the heat will accumulate from the subsequent pulses in the focal volume and elevate the target temperature on the surface and in the bulk. The characteristic thermal diffusion time in glass is about 1 μs for a volume of 0.3 μm3. This thermal diffusion time will vary from glass-to-glass according to their composition. However for this report, we are taking this value as a reference. In comparison to this thermal diffusion time, the separation time between two pulses is much smaller; 77, 125, and 250 ns for 13-, 8-, and 4-MHz repetition rates, respectively. Even though all the aforementioned times are much less than the heat diffusion time of 1 μs, the heat accumulation will be high in and around the focal volume at higher repetition rate compared to lower repetition rate. As a result, the energy per pulse required to start the breakdown reduces as the pulse repetition rate is increased. This breakdown threshold energy per pulse is found to be 2.032, 1.338, and 0.862 μJ for 4, 8, and 13 MHz, respectively.
Effect of dwell time
Effect of laser polarization
All the experiments discussed above were performed by circular polarization of femtosecond laser pulses. We also wanted to investigate whether the linear polarization changes the growth mechanism of nanostructures on the laser-irradiated target glass. The effect of laser polarization on the ablation of various materials has been studied by many researchers. Hee et al. studied the effect of polarized femtosecond laser pulses on the generation of relief gratings on (111) silicon substrate using a novel interferometer . The ablation was performed by focusing two interfering femtosecond laser beams under different polarization combinations. In their investigation, they found that p:-p-polarization has the lowest ablation threshold and generates the deepest grating depth among other polarization combinations (s-:s-polarization; c-:c-polarization). Camacho-Lopez et al. investigated the growth of grating-like structures on titanium films by circular (c-) and linear (p-) polarizations . They discovered that there was no formation of grating-like structures when the substrate was irradiated with circularly polarized light. However, when linearly polarized laser pulses were utilized, the grating-like structures were generated at the fluence well below the ablation threshold for the titanium film. Furthermore, Venkatakrishnan et al. also found in their study of polarization effects on ultrashort-pulsed laser ablation of thin metal films that linear (p-) polarization has an ablation threshold less than that for circular polarization .
Looking at the SEM images in Figure 10, these changes can be better understood. Figure 10a shows the SEM image of the target irradiated with circularly polarized laser pulses with 4-MHz repetition rate at the dwell time of 0.25 ms. It can be seen that there is no evident of tip growth most likely due to the inadequate ablated material into the plasma. When the target was irradiated with linearly (p-) polarized pulses with the same laser parameters, as depicted in Figure 10d, a high number of nanotips were found to be growing on the target surface. This is only possible if the linearly polarized pulses ablated an adequate amount of material from the target into the plasma in order for the nanotips to initiate and complete their growth. To further make sure if this is the case for other laser parameters with linear polarization, we also irradiated targets at 0.5-ms dwell time for 4 MHz and at 0.25 ms for 8 MHz. The corresponding SEM images of these experiments are shown in Figure 10. For each parameter, it was found that the growth of nanotips improved in terms of density of nanotips over large target surface at each parameter. From this result, it can be understood that the linear (p-) polarization does not really alter the nanotip growth mechanism but rather it enhances it. Since linearly polarized pulses ablate material more effectively even at the same pulse energy in comparison to circular polarization, it will take fewer numbers of pulses while using linear polarization to reach each growth stage explained in Figure 8. Now that we know how the growth of nanotips is affected using various femtosecond laser parameters, it will be beneficial to perform in situ analysis of the plasma expansion, the process temperature, and pressure gradient for each combination of the laser parameters. This future work will help us find out the exact combination of femtosecond laser parameters which will produce more uniform and maximum number of nanotips over the large surface of the dielectric targets.
In summary, we have discussed the growth of leaf-like nanostructures with nanoscale apex from dielectric target material by femtosecond laser irradiation at megahertz pulse repetition rates. In our synthesis method, the whole growth process occurs in an open air at ambient conditions in the presence of nitrogen gas flow without the use of any catalyst. The dielectric target provides two roles: first as the source for building material and second as the substrate upon which these leaf-like nanotips can grow.
The growth mechanism of nanotips is explained by classic thermal diffusion. We observed the growth of individual and multiple nanotips from relatively small single droplets at shorter pulse width; whereas when the pulse width was increased, the nanotips grew mainly from the film of the molten target material and the large deposited droplets of molten material.
The laser specifications (laser pulse width, pulse repetition rate, and laser polarization), processing parameters (dwell time), and gas flow rate control the number of tips synthesized and, to some extent, the size of tips. In our investigation, we found the clear transformation of the kind of nanotips that grow under various conditions. In further experiments, we found that for a given dwell time, the number of nanotips that grow on target surface increases with increasing pulse repetition rate. However, this was only observed for certain dwell times. If the laser machining was continued on a spot beyond a certain dwell time for each repetition rate, we observed fewer nanotips grown on the target along with the growth of increasing number of micronanoparticles and remelting of previously deposited plasma material. The dwell time was observed to be influencing the nanotip growth in a similar manner as pulse repetition rate; at low dwell time, only the growth of a small number of stems was observed. As the dwell time was increased for a given repetition rate, an increasing number of stems and nanotips were found to be growing on the irradiated target surface. Finally, we studied the effect of linear polarization on the growth of leaf-like nanotips. We observed the enhanced number of nanotips grown on the target surface in comparison to machining under circular polarization of the laser for the same given laser parameters.
Future work will involve the in situ analysis of plasma interactions with nitrogen gas flow and incoming laser pulses, the pressure and the temperature gradient of target surface, and the expanding plasma. Understanding the aforementioned phenomena in situ will provide more control and help us grow more uniform nanotips over the large surface area of the target. This study was carried out with silicon substrate, but we believe that other semiconductor materials may also generate similar phenomena.
NP was a candidate of Master of Applied Science. KV is the co-supervisor of NP. BT is the supervisor of NP.
This research is funded by the Natural Science and Engineering Research Council of Canada and Ministry of Research and Innovation, Ontario, Canada.
- Levchenko I, Ostrikov K, Long JD, Xu S: Plasma-assisted self-sharpening of platelet-structures single-crystalline carbon nanocones. Appl Phys Lett 2007, 91: 113115. 10.1063/1.2784932View ArticleGoogle Scholar
- Liu C, Hu Z, Wu Q, Wang X, Chen Y, Sang H, Zhu J, Deng S, Xu N: Vapor-solid growth and characterization of aluminum nitride nanocones. J Am Chem Soc 2005, 127: 1318–1322. 10.1021/ja045682vView ArticleGoogle Scholar
- Cheng C-L, Chao S-H, Chen Y-F: Enhancement of field emission in nanotip-decorated ZnO nanobottles. J Cryst Growth 2009, 311: 381–4384.Google Scholar
- Chen H, Pasquier AD, Saraf G, Zhong J, Lu Y: Dye-sensitized solar cells using ZnO nanotips and Ga-doped ZnO films. Semicond Sci Technol 2008, 23: 045004. 10.1088/0268-1242/23/4/045004View ArticleGoogle Scholar
- Li YB, Bando Y, Golberd D: ZnO nanoneedles with tip surface perturbations: excellent field emitters. Appl Phys Lett 2004, 83: 3603–3605.View ArticleGoogle Scholar
- Shen G, Bando Y, Liu B, Goldberg D, Lee C-J: Characterization and field-emission properties of vertically aligned ZnO nanonails and nanopencils fabricated by a modified thermal-evaporation process. Adv Funct Mater 2006, 16: 410–416. 10.1002/adfm.200500571View ArticleGoogle Scholar
- Lo HC, Das D, Hwang JS, Chen KH, Hsu CH, Chen CF, Chen LC: SiC-capped nanotips arrays for field emission with ultralow turn-on field. Appl Phys Lett 2003, 83: 1420–1422. 10.1063/1.1599967View ArticleGoogle Scholar
- Yao I-C, Lin P, Tseng T-Y: Nanotip fabrication of zinc oxide nanorods and their enhanced field emission properties. Nanotechnology 2009, 20: 125202. 10.1088/0957-4484/20/12/125202View ArticleGoogle Scholar
- Moening JP, Thanawala SS, Georgiev DG: Formation of high-aspect-ratio protrusions on gold films by localized pulsed laser irradiation. Appl Phys A: Mater Sci Process 2009, 95: 635–638. 10.1007/s00339-009-5166-6View ArticleGoogle Scholar
- Moening JP, Georgiev DG, Lawrence JG: Focused ion beam and electron microscopy characterization of nanosharp tips and microbumps on silicon and metal thin films formed via localized single-pulse laser irradiation. J Appl Phys 2011, 109: 014304. 10.1063/1.3524367View ArticleGoogle Scholar
- Kuznetsov AI, Koch J, Chichkov BN: Nanostructuring of thin gold films by femtosecond lasers. Appl Phys A: Mater. Sci Process 2009, 94: 221–230. 10.1007/s00339-008-4859-6View ArticleGoogle Scholar
- Kuznetsov AI, Unger C, Koch J, Chichkov BN: Laser-induced jet formation and droplet ejection from thin metal films. Appl Phys A: Mater Sci Process 2012, 106: 479–487. 10.1007/s00339-011-6747-8View ArticleGoogle Scholar
- Her T-H, Finlay RJ, Wu C, Deliwala S, Mazur E: Microstructuring of silicon with femtosecond laser pulses. Appl Phys Lett 1998, 73: 1673–1675. 10.1063/1.122241View ArticleGoogle Scholar
- Shen MY, Crouch CH, Carey JE, Mazur E: Femtosecond laser-induced formation of submicrometer spikes on silicon in water. Appl Phys Lett 2004, 85: 5694–5696. 10.1063/1.1828575View ArticleGoogle Scholar
- Sivakumar M, Venkatakrishnan K, Tan B: Synthesis of nanoscale tips using femtosecond laser radiation under ambient condition. Nanoscale Res Lett 2010, 5: 438–441. 10.1007/s11671-009-9502-8View ArticleGoogle Scholar
- Tan B, Panchatsharam S, Venkatakrishnan K: High rep-rate femtosecond laser forming sub-10 μm interconnection vias. J Phys D: Appl Phys 2009, 42: 065102. 10.1088/0022-3727/42/6/065102View ArticleGoogle Scholar
- Fan C-H, Longtin JP: Modelling optical breakdown in dielectrics during ultrafast laser processing. Appl Opt 2001, 40: 3124–3131. 10.1364/AO.40.003124View ArticleGoogle Scholar
- Varel H, Ashkenasi D, Rosenfeld A, Herrmann R, Noack F, Campbell EEB: Laser-induced damage in SiO2 and CaF2 with picosecond and femtosecond laser pulses. Appl Phys A: Mater Sci Process 1996, 62: 293–294. 10.1007/BF01575098View ArticleGoogle Scholar
- Sanner N, Uteza O, Bussiere B, Coustillier G, Leray A, Itina T, Sentis M: Measurement of femtosecond laser-induced damage and ablation thresholds in dielectrics. Appl Phys A: Mater Sci Process 2009, 94: 889–897. 10.1007/s00339-009-5077-6View ArticleGoogle Scholar
- Du D, Liu X, Korn G, Squier J, Mourou G: Laser-induced breakdown by impact ionization in SiO2 with pulse widths from 7 ns to 150 fs. Appl Phys Lett 1994, 64: 3071–3073. 10.1063/1.111350View ArticleGoogle Scholar
- Stuart BC, Feit MD, Rubenchik AM, Shore BW, Perry MD: Laser-induced damage in dielectrics with nanosecond to subpicosecond pulses. Phys Rev Lett 1995, 74: 2248–2251. 10.1103/PhysRevLett.74.2248View ArticleGoogle Scholar
- Dausinger F, Lubatschowski H, Lichtner F: Femtosecond technology for technical and medical applications. Top Appl Phys 2004, 96: 75–91. 10.1007/978-3-540-39848-6_6View ArticleGoogle Scholar
- Schaffer CB, Garcia JF, Mazur E: Bulk heating of transparent materials using a high repetition-rate femtosecond laser. Appl Phys A: Mater Sci Process 2003, 76: 351–354. 10.1007/s00339-002-1819-4View ArticleGoogle Scholar
- Hee CW, Ngoi BKA, Lim LEN, Venkatakrishnan K, Liang WL: Effect of polarization on femtosecond pulsed laser ablation on surface relief gratings. Opt Laser Technol 2005, 37: 93–98. 10.1016/j.optlastec.2004.02.016View ArticleGoogle Scholar
- Camacho-Lopez S, Evans R, Escobar-Alarcon L, Camacho-Lopez MA: Polarization-dependent single-beam laser-induced grating-like effects on titanium films. Appl Surf Sci 2008, 255: 3028–3032. 10.1016/j.apsusc.2008.08.085View ArticleGoogle Scholar
- Venkatakrishnan K, Tan B, Stanley P, Sivakumar NR: The effect of polarization on ablation of thin metal films using ultra short pulsed laser. J Appl Phys 2002, 92: 1604. 10.1063/1.1487453View ArticleGoogle Scholar
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.