In this work we present the refractive index sensing performance comparison between resonant (R-NPs) and silicon dioxide (SiO2-NPs) high-aspect ratio nano-pillars arrays. Both arrays have been fabricated by laser interference lithography and reactive ion etching. The R-NPs are made by a multilayer of silicon oxide and silicon nitride distributed to act as a vertical resonant cavity with two Bragg reflectors. Several chips containing eight periodic arrays of R-NPs and SiO2-NPs were implemented following the presented fabrication process, having a height in the order of 2.5 μm, a diameter in the order of 200 nm, different pitches and aspect ratio up to 9.8. Finally, the optical responses of these arrays were measured by infiltration of fluids with different refractive indexes. The main conclusion is that sensitivity obtained for the R-NPs is more than two times higher in comparison with the SiO2-NPs sensitivity (3724 cm−1/RIU and 1652 cm−1/RIU, respectively).
© 2016 Optical Society of America
Recent works in the literature use sub-micrometric and even nanometric pillars for chemical and biochemical sensing [1–3], generally conforming periodic arrays over a surface. The use of this typology of surfaces for sensing has proven several advantages, since the sensing area is remarkably increased compared with a flat surface and, on the other hand, there are also interesting confinement effects due to the proximity of the pillars. Usually, the optical reflectance or transmittance of the array is measured as a function of the refractive index of the medium (refractive index sensor), also named as bulk sensing. The optical response of these sensors mainly depends on the geometrical parameters of the pillars and the arrays, the optical properties of the materials used and the distribution of materials (film stack). The developments in micro-nano fabrication techniques have allowed fabricating not only pillars of a single material, but also vertical Fabry-Perot resonators consisting of a pair of Bragg reflectors and a central cavity, with micrometric and even sub-micrometric sizes. The use of resonators of this range of sizes has several applications from the photonic point of view, mainly in obtaining high Q-factor modes [4–6].
In this work we compare the refractive index sensing performance and the fabrication process of two different typologies of nano-pillars (NPs) arrays. The first one is based on silicon dioxide NPs, with heights up to 2.9 μm, diameters in the order of 180-300 nm, and pitches down to 300 nm. The second one based on arrays of resonant nanopillars (R-NPs), built each NP as a Fabry-Perot resonator formed by a multilayer of SiO2 and Si3N4, distributed as a micro-cavity and two Bragg reflectors over a glass substrate. These R-NPs have the same dimensions than the SiO2-NPs. The R-NPs combine the advantages of sub-micrometric size of the pillars and the micro-resonators from the sensing and photonic point of view. Recently we presented R-NPs arrays as refractive index sensor, which allows reaching a value of limit of detection for refractive index sensing in the order of 10−5 RIU (refractive index units) .
For the comparison, it is described the fabrication process developed for obtaining the SiO2-NPs and the R-NPs arrays, which combines laser interference lithography (LIL) with several processes of pattern transfer such us lift-off and reactive ion etching (RIE). There are numerous studies reporting fabrication processes of pillars with high Aspect Ratio (AR) [8–13], and also of micro and nanopillars [1,14]. However, to the best of our knowledge, there are not published works about the fabrication to obtain periodic arrays  of R-NPs with the dimensions and materials reported in this work by using LIL. Moreover, compared to other techniques, LIL allows fabricating larger areas in shorter time than e-beam lithography [1,10], and provides a lower density of defects than self-assembly or nano-sphere lithography [8,11,12]. On the other hand, some of these works have studied the formation of high AR silicon pillars using RIE [8,10], and we have also reported the fabrication of silicon pillars using LIL and RIE [16,17].
Finally we have characterized the optical response of both configurations of pillars by infiltrating liquids with different refractive indexes, obtaining values for the refractive index sensitivity of 3724 cm−1/RIU for the R-NPs and 1653 cm−1/RIU for the SiO2 NPs.
2.1 Fabrication of SiO2–NPs and R-NPs arrays
Glass substrates are used for the chips with a thick SiO2 layer deposited on top by e-beam evaporation in the case of SiO2-NPs; or with a SiO2/Si3N4 multilayer stack deposited by chemical vapor deposition in the case of R-NPs. The multilayer consists of two Bragg reflectors with 5 or 9 bilayers of 110 nm SiO2 and 80 nm Si3N4, and a central cavity of 210 nm of SiO2. A spin coater (WS-650S-6NPP, Laurell) is used for the deposition of an anti-reflective coating (ARC, XHRiC-16 from Brewer Science) and a photoresist (NR7 from Futurrex). An e-beam evaporator (Classic 5000, Pfeiffer) is used for the deposition of the needed films of SiO2 and chromium (Cr). The pattern is created by using a Lloyd’s mirror LIL set-up with a He-Cd laser (325 nm), and is transferred by means of RIE (Plasmalab 80, Oxford Instruments) using oxygen and fluorine gases like CHF3, CF4 and SF6. Finally, the remaining Cr after the pattern transfer is removed with a commercial Cr etchant (Sigma-Aldrich). The final fabricated nano-pillars as well as all the previous steps and the etching optimization experiments have been morphologically characterized with a Scanning Electron Microscope (SEM, Ultraplus Carl Zeiss).
The fabrication processes of SiO2-NPs and R-NPs start with the deposition on the substrates of a SiO2 layer or a multilayer of SiO2 and Si3N4 respectively (see Fig. 1(e) for comparison of both NPs). For the first case, a 3 μm thick SiO2 layer is deposited by e-beam evaporation while, in the second case, substrates are acquired with the multilayer already prepared. The following steps are the same for both fabrication processes, which continue with the preparation of an adequate stack for the LIL on top of the SiO2 layer or the multilayer of SiO2 and Si3N4. This stack is composed by an anti-reflective coating (ARC), a negative photoresist layer (PRL) and a SiO2 interlayer , necessary for a vertical pattern transference and also because it provides a suitable profile for lift-off. A pattern of circular holes distributed in a square lattice is created onto the PRL by exposing samples twice in a Lloyd’s mirror LIL set-up . The exposure dose is adjusted to obtain holes with diameters between 150 nm and 250 nm (Fig. 1(a)). The pattern is then transferred from the PRL to the SiO2 interlayer and from the SiO2 to the ARC by means of CF4 and O2 RIEs respectively (Fig. 1(b)). The adequate undercutting for the lift-off can be obtained in this step with a little bit of O2 overetch. A Cr layer with a thickness of 150 nm is deposited by e-beam evaporation on this SiO2/ARC pattern, and samples are immersed then in a solution called SC1 of RCA . This lift-off process leaves a pattern of Cr pillars on the surface (Fig. 1(c)), which is used as hard mask for the RIE of the nanopillars. The Cr Mask is removed then by using a commercial Cr etchant (Fig. 1(d)).
2.2 Refractive index sensing assay
For characterizing the pillars as refractive index sensor, we infiltrate the different fluids by immersion the topside of the sample where the NPs arrays are located with a holder manufactured ad hoc with the pillars facing down. The optical interrogation was performed vertically from the backside of the transparent substrate . This optical interrogation was carried out by Fourier Transform Visible Infrared Spectroscopy (FT-VIS-IR) at a normal grazing angle; the resolution employed was of 4 cm−1 with 500 scans to improve the signal to noise ratio. For each Refractive Index (RI) media given for the different fluids infiltrated, we obtained the optical response of the NPs arrays at the visible spectral range. We intended to identify the narrowest resonances to achieve a higher accuracy in the wavelength position along the experiment. We used increasing concentrations of ethanol in water, from pure water to 10%, 20% and 100% of ethanol. The RIs of the water–ethanol mixtures as a function of ethanol concentration were determined from P. Liebetraut et al .
3. Results and discussion
3.1. Fabrication of the high AR NPs
Several parameters of the fabrication process were optimized in order to achieve pillars with high aspect ratio and low pitch. RIE gases used and the rest of RIE parameters were also optimized. The main problem found was the central overetch of the pillars during the process due to different interactions among ions, radicals, electrons and volatile molecules, or with pillars edges when they are tall and close to each other . If this central overetch is too big, the pillars can collapse or even be broken at the end of the process. The closer the pillars are to each other (i.e. the smaller the pitch is), the higher incidence these interactions have in the etching. As this overetch has been found unavoidable for the high AR sought, the maximum achievable height of the pillars is higher for 600 nm pitch than for 400 nm and 300 nm pitch.
Another issue to take into consideration when fabricating high AR NPs is the sidewall tapering during the etching. The combination of high AR with features that get closer and closer during the etching can decrease the etch rate drastically and even stop the etching if the bases touch each other [22,23].
The third parameter that needs to be minimized during these etching processes is the sidewall roughness of the NPs. Figure 2(a) shows three SEM images with examples of these three undesired effects that have to be minimized during the optimization process.
The optimization started with etchings based on the gases CF4, a mixture of CHF3/SF6 and a mixture of CHF3/O2, using different values of chamber pressure, radio-frequency (RF) power and gases concentration. After every etching, the obtained pillars were characterized by SEM. The central overetch of the pillars, their tapering and the roughness of their surface were evaluated from 1 (worst) to 5 (best) points in order to have a numerical value that will help choosing the best results. The first batch of experiments shows that the mixture of CHF3/O2 gave the best results. Figure 2(b) shows a graph where it can be seen at the left half the evaluation of these three processes that use CF4, CHF3/SF6 and CHF3/O2 (A, B and C respectively), with 10 mT of chamber pressure and 50 W of RF power. The best results were obtained using the mixture of CHF3/O2.
Once selected the best gas mixture, the rest of etching parameters were optimized. The chamber pressure was varied between 10 mT and 100 mT, the RF power between 25 W and 150 W, and the O2 concentration in the mixture of CHF3/O2 between 3% and 14%. It was found that by using 50 mT of chamber pressure, the three characteristics improved (process D). Besides, the central overetch was minimized using 75 W of RF power (process E). Finally, the 6% O2 concentration gave the best results in the evaluated range. After the numerical evaluation of all the processes, the optimum RIE recipe found is the process F of the Fig. 2(b), which uses a mixture of CHF3/O2 (with 6% of O2), 50 mT of chamber pressure and 75 W of RF power. With this recipe we have fabricated several chips of SiO2-NP with pitches of 300, 400 and 600 nm, reaching a height of 2.9 μm in the case of 600 nm of pitch (Fig. 3(a)), and an AR of 9.8 in the case of 400 nm of pitch (Fig. 3(b)).
The adaptation of the fabrication process to substrates with the SiO2/Si3N4 multilayer stack was quite direct because we found that the vertical etch rate for SiO2 using this RIE recipe was 17 nm/min, and 14.5 nm/min for Si3N4. Hence, the average etch rate for the multilayer stack was 16.1 nm/min. Thanks to this small difference, the height of the samples with the multilayer stack is very similar to the ones obtained with the SiO2 film. Besides, the central overetch, the tapering and the roughness do not suffer important changes when etching the multilayer stack. In addition, as can be seen in the SEM images of Fig. 4(a), the different lateral etching of SiO2 and Si3N4 is negligible in the fabricated samples.
Thus, using the optimized RIE process (F in Fig. 3(b)), we successfully fabricated several R-NP arrays with sizes of 20 mm x 10 mm. Two different stack configurations and the three pitches were studied, reaching also AR values up to 9.8. Their characteristics are summarized in the table of Fig. 4(a). Chips have eight arrays of R-NPs as it can be seen in Fig. 4(b), and it was obtained by using a Kapton mask during the Cr evaporation for the lift-off process.
3.2. Bulk sensing performance of SiO2-NPs versus R-NPs
The main focus of the present work is to characterize and compare the yield of the two different sorts of nano-pillars above mentioned (SiO2-NPs and R-NPs). For that purpose we measured the reflection spectra as a function of wavenumber of both types of NPs arrays fabricated with the same pitch (300 nm for this case), and for different RI fluids. Figure 5 shows the reflection spectra of NPs. It is remarkable the difference in the amplitude of the reflectivity signal of both types of NPs. In addition it can be observed sharper mode for the case of the R-NPs.
The wavenumber position of the resonance depends on the height, diameter and lattice parameter for the SiO2-NPs; and on the diameter, lattice parameter and the thickness of the cavity layer for the R-NPs. The SiO2-NPs spectra have several dips and peaks than can be monitorized, whereas the R-NPs have a single resonance dip. The position of this dip varies when the fabrication parameters change, as can be seen in Fig. 6 and Fig. 5(b). For a pitch of 300 nm the dip is centered around 17250 cm−1, whereas for 400 and 600 nm is centered around 17900 and 17500 cm−1, respectively.
The NPs arrays were infiltrated with different Di-H2O/ET-OH fraction solutions and subsequently optically interrogated in order to monitor the shift of the narrowest resonance position. The shift can be expressed as a function of RI (Fig. 7). The ethanol content of the solutions was: 10, 20 and 100% (vol.). Deionized water was employed as the reference.
Because the R-NPs arrays are also mainly based on SiO2, we found that there were not significant differences in their wettability, and therefore we did not observed any difference during the infiltration of the different RI fluids between both types of arrays.
However, we observed that the SiO2-NPs reflectivity signal was much lower than for the case of the R-NPs
As a result, we observed that sensitivity, calculated as the wavenumber shift divided by the refractive index variation (cm−1/RIU), reached a value of 3724.48 cm−1/RIU for the R-NPs in contrast to the sensitivity obtained for the SiO2-NPs that reached a value of 1652.94 cm−1/RIU. Both bulk sensing responses were obtained exactly under the same conditions: the same ethanol-water solutions, humidity and room temperature at 20 °C. The obtained experimental uncertainty, considering the resolution of the system employed of 4 cm−1 for a coverage factor of k = 3, is of 4.9 cm−1 for the SiO2-NPs, and 3.7 cm−1 for the R-NPs, which results in a limit of detection (LoD) of 2.9x10−3 RIU and 9.9x10−4 RIU, respectively. However, if we only consider as uncertainty factor the maximum resolution of our experimental system, which is 1 cm−1, this LoD is of 2.6x10−4 in comparison with 6.04x104.
We have also measured the sensitivity for the three different values of pitch of the R-NPs fabricated. For a pitch of 400 nm the obtained sensitivity is of 4068.39 cm−1/RIU, whereas for 600 nm this value is of 2303.33 cm−1/RIU. As a general rule, lower values of pitch result in a better value of sensitivity. This is clear in the case of 600 nm; however, values for 300 and 400 nm are in the same order. This is probably due to the different values of diameter and height for the two different pitch arrays.
In this paper we validate the feasibility to use laser interference lithography (LIL) and reactive ion etching (RIE) to successfully fabricate large arrays of high aspect ratio SiO2-NPs and SiO2/Si3N4 R-NPs. The fabrication processes parameters for both sensing arrays are similar, being the main different the vertical RIE etch rate, resulted in 17 nm/min for SiO2 and 16.1 nm/min for the multilayer of SiO2/Si3N4.
Once the fabrication process to obtain SiO2/Si3N4 R-NPs arrays resulted not to be of a high complexity in comparison with the SiO2-NPs arrays, with no differences in their wettability response, in this paper we determine for the first time the benefits of using R-NPs in comparison with SiO2-NPs arrays in terms of bulk sensitivity. In fact we demonstrate that chemical bulk sensitivity of R-NPs arrays doubles the SiO2-NPs and the LoD is also improved by a factor of 4, demonstrating their higher performance and its suitability to be used as biochemical transducers.
This work has been funded by the projects ENVIGUARD, founded by the European Commission (OCEAN-2013-614057) and the Spanish Ministry “Ministerio de Economía y Competitividad” under the projects PLATON (Ref.:TEC2012-31145).
References and links
1. M. Holgado, C. A. Barrios, F. J. Ortega, F. J. Sanza, R. Casquel, M. F. Laguna, M. J. Bañuls, D. López-Romero, R. Puchades, and A. Maquieira, “Label-free biosensing by means of periodic lattices of high aspect ratio SU-8 nano-pillars,” Biosens. Bioelectron. 25(12), 2553–2558 (2010). [CrossRef] [PubMed]
3. B. Choudhury, R. Casquel, M. J. Bañuls, F. J. Sanza, M. F. Laguna, M. Holgado, R. Puchades, A. Maquieira, C. A. Barrios, and S. Anand, “Silicon nanopillar arrays with SiO2 overlayer for biosensing application,” Opt. Mater. Express 4(7), 1345–1354 (2014). [CrossRef]
4. D. M. Whittaker, P. S. Guimaraes, D. Sanvitto, H. Vinck, S. Lam, A. Daraei, J. A. Timpson, A. M. Fox, M. S. Skolnick, Y.-L. D. Ho, J. G. Rarity, M. Hopkinson, and A. Tahraoui, “High Q modes in elliptical microcavity pillars,” Appl. Phys. Lett. 90(16), 161105 (2007). [CrossRef]
5. V. N. Astratov, S. Yang, S. Lam, B. D. Jones, D. Sanvitto, D. M. Whittaker, A. M. Fox, M. S. Skolnick, A. Tahraoui, P. W. Fry, and M. Hopkinson, “Whispering gallery resonances in semiconductor micropillars,” Appl. Phys. Lett. 91(7), 071115 (2007). [CrossRef]
6. J. P. Reithmaier, G. Sek, A. Löffler, C. Hofmann, S. Kuhn, S. Reitzenstein, L. V. Keldysh, V. D. Kulakovskii, T. L. Reinecke, and A. Forchel, “Strong coupling in a single quantum dot-semiconductor microcavity system,” Nature 432(7014), 197–200 (2004). [CrossRef] [PubMed]
7. A. L. Hernández, R. Casquel, M. Holgado, I. Cornago, F. J. Sanza, B. Santamaría, M. Maigler, F. Fernández, A. Lavín, and M. F. Laguna, “Arrays of resonant nanopillars for biochemical sensing,” Opt. Lett. 40(10), 2370–2372 (2015). [CrossRef] [PubMed]
8. T. Trifonov, A. Rodríguez, F. Servera, L. F. Marsal, J. Pallares, and R. Alcubilla, “High‐aspect‐ratio silicon dioxide pillars,” Phys. Status Solidi., A Appl. Mater. Sci. 202(8), 1634–1638 (2005). [CrossRef]
9. V. Ovchinnikov, A. Malinin, S. Novikov, and C. Tuovinen, “Fabrication of silicon nanopillars using self-organized gold–chromium mask,” Mater. Sci. Eng. B 69–70, 459–463 (2000). [CrossRef]
10. Y. F. Chang, Q. R. Chou, J. Y. Lin, and C. H. Lee, “Fabrication of high-aspect-ratio silicon nanopillar arrays with the conventional reactive ion etching technique,” Appl. Phys., A Mater. Sci. Process. 86(2), 193–196 (2006). [CrossRef]
11. C. L. Cheung, R. J. Nikolić, C. E. Reinhardt, and T. F. Wang, “Fabrication of nanopillars by nanosphere lithography,” Nanotechnology 17(5), 1339–1343 (2006). [CrossRef]
12. L. Sainiemi, H. Keskinen, M. Aromaa, L. Luosujärvi, K. Grigoras, T. Kotiaho, J. M. Mäkelä, and S. Franssila, “Rapid fabrication of high aspect ratio silicon nanopillars for chemical analysis,” Nanotechnology 18(50), 505303 (2007). [CrossRef]
13. K. J. Morton, G. Nieberg, S. Bai, and S. Y. Chou, “Wafer-scale patterning of sub-40 nm diameter and high aspect ratio (>50:1) silicon pillar arrays by nanoimprint and etching,” Nanotechnology 19(34), 345301 (2008). [CrossRef] [PubMed]
14. H. S. Wasisto, S. Merzsch, A. Stranz, A. Waag, E. Uhde, T. Salthammer, and E. Peiner, “Silicon resonant nanopillar sensors for airborne titanium dioxide engineered nanoparticle mass detection,” Sens. Actuators B Chem. 189, 146–156 (2013). [CrossRef]
15. H. I. Smith, “Low cost nanolithography with nanoaccuracy,” Phys. E 11(2-3), 104–109 (2001). [CrossRef]
16. S. Domínguez, I. Cornago, O. García, M. Ezquer, M. J. Rodríguez, A. R. Lagunas, J. Pérez-Conde, and J. Bravo, “Design, optimization and fabrication of 2D photonic crystals for solar cells,” Phot. Nano. Fund. Appl. 11(1), 29–36 (2013). [CrossRef]
17. I. Cornago, S. Domínguez, M. Ezquer, M. J. Rodríguez, A. R. Lagunas, J. Pérez-Conde, R. Rodriguez, and J. Bravo, “Periodic nanostructures on Unpolished substrates and their integration in solar cells,” Nanotechnology 26(9), 095301 (2015). [CrossRef] [PubMed]
18. T. A. Savas, M. L. Schattenburg, J. M. Carter, and H. I. Smith, “Large‐area achromatic interferometric lithography for 100 nm period gratings and grids,” J. Vac. Sci. Technol. B 14(6), 4167–4170 (1996). [CrossRef]
19. W. Kern, “The evolution of silicon wafer cleaning technology,” J. Electrochem. Soc. 137(6), 1887–1892 (1990). [CrossRef]
21. R. A. Gottscho, C. W. Jurgensen, and D. J. Vitkavage, “Microscopic uniformity in plasma etching,” J. Vac. Sci. Technol. B 10(5), 2133–2147 (1992). [CrossRef]
22. J. Kiihamäki, “Deceleration of silicon etch rate at high aspect ratios,” J. Vac. Sci. Technol. A 18(4), 1385–1389 (2000). [CrossRef]
23. I. Cornago, H. J. Choi, J. G. Kim, and G. Barbastathis, “Simple and Low-Cost Fabrication Process of High Aspect Ratio Periodic Nanocone Structures,” presented at the 38th International Conference of Micro and Nano Engineering, Toulouse, France, 16–20 Sept. 2012.