Wafer-scale fabrication of 3nm-thick silicon nitride nanopore membranes

Solid-state nanopores for use in DNA sequencing are fabricated using a polycrystalline silicon sacrificial layer process.
17 February 2016
Itaru Yanagi

Nanopore sequencing (which refers to threading DNA through nanometer-sized pores) is a promising approach for achieving long-read, label-free, single-molecule DNA sequencing with very high throughput and low cost.1, 2 Personalized medicine is expected to be a growth area in the future, and would require fast DNA sequencing.3 Therefore, both biological and solid-state nanopores have been intensively studied in recent years. The most used DNA sequencing method, common to both biological and solid-state nanopores, consists of detecting changes in ionic current through a nanopore during the translocation of DNA and identifying the four types of nucleotides through these changes.

The spatial resolution of a nanopore sensor is determined by the thickness of the membrane. As the distance between neighboring nucleotides in DNA is very short (approximately 0.4nm), thinning the membrane is a very important issue for highly accurate discrimination of each nucleotide in DNA.

Here, we report stable wafer-scale fabrication of ultra-thin silicon nitride membranes using the polycrystalline silicon (poly-Si) sacrificial process. The process flow for fabricating the membranes is shown in Figure 1. The most important part of this process is wet etching of the Si substrate and the poly-Si sacrificial layer. Tetramethylammonium hydroxide and potassium hydroxide (KOH) aqueous solutions do not etch the silicon nitride (Si3N4) film, due to its strong chemical resistance to these solutions. In addition, this process has the advantage of enabling the fabrication of membranes with a small circular area approximately 500nm in diameter that reduces the probability of initial breakage of the membrane. Figure 2 presents detailed information on the thickness of the deposited Si3N4 films measured using ellipsometry. The film thickness at each point on the eight-inch wafer is shown in Figure 2(a), and the cumulative probability of the thickness value is shown in Figure 2(b). The variation in the film thickness was quite small (3.10–3.35nm), and the average thickness was 3.18nm. Figure 3(a) presents cross-sectional scanning transmission electron microscope (STEM) images of the Si3N4 films at three points on the wafer: (1), (2), and (3) in Figure 3(a). From these STEM images, the Si3N4 film thickness was found to be approximately 2.7nm, which agrees with the thickness measured using ellipsometry. Top-view transmission electron microscopy (TEM) images of the membrane are shown in Figure 3(b). This figure confirms that the poly-Si sacrificial layer can be removed by etching with aqueous KOH solution, and that clean Si3N4 membranes can be fabricated. The results presented in Figures 2 and 3 confirm that Si3N4 membranes, approximately 3nm thick, can be fabricated using the poly-Si sacrificial layer process.

Figure 1. Process flow for the fabrication of membranes using the poly-crystalline silicon (poly-Si) sacrificial layer process. Wet etching performed with tetramethylammonium hydroxide (TMAH) and potassium hydroxide (KOH). Si3N4: Silicon nitride. Si sub: Silicon substrate.

Figure 2. Thickness of SiN films measured using ellipsometry. (a) The thickness measured at 25 points spread over the wafer. The reflective index was set to 2.0. (b) Cumulative probability of the measured thickness values.

Figure 3. Cross-sectional scanning tunneling electron microscope (STEM) images of the SiN layer and top views of the SiN membrane. (a) The SiN layer at three different points on the wafer was observed at 2000k-fold magnification. The wafer is slightly thicker toward the edges. (b) A transmission electron microscopy (TEM) image of the entire membrane at 20k-fold magnification. A magnified TEM image at 100k-fold magnification.

Figure 4(a) presents TEM images of nanopores (with diameters approximately 2–6nm) fabricated via focused-electron-beam etching. The mean diameter (φm) was defined, using an ellipsoidal approximation, as

where φl and φs are the major and minor axes, respectively, of the nanopore measured from the TEM image. The relationship between φm and the conductance of the ionic current through the nanopore (G0) is illustrated in Figure 4(b). The currents were recorded 3–5s after a voltage (0.1V) was applied. The plotted measurements were fitted with the theoretically calculated lines,4 using the current through the nanopore and voltage across the membrane. The diameter of the nanopore was controlled by adjusting the irradiation time and the electron flux of the focused-electron beam.

Figure 4. TEM images of nanopores and relationship between the ionic current through nanopores and their diameter. (a) TEM images of fabricated nanopores at 400k-fold magnification. Each left image shows the raw image of each right image. (b) Relationship between ionic conductance through nanopores G0)and their diameters (φm). A total of 28 points are plotted within φm=1.65–6.14nm.

Figure 5 shows a scatter plot and histogram of the ionic-current blockades during double-stranded DNA translocations through the nanopore with a φm of 2.88nm. The voltage applied was 0.3V. The mean conductance blockade, ΔGP, is approximately 9.7nS. We used a geometric model5 to estimate ΔGP from the conductance measurements of the inside of the nanopore and the access resistance region during DNA translocation. The estimated ΔG is 9.73nS. This value is in good agreement with the experimental results.

Figure 5. Ionic current through a nanopore shows current-blockade events, indicating DNA translocations through the nanopore. Left: Time traces of ionic current, under an applied voltage of 0.3V, through a nanopore with a φmof 2.88nm. Second panel: A magnified view of a current-time trace that shows typical ionic-current-blockade events. Right: Scatter plot and histogram of the current-blockade events. ΔIp: Mean ionic current blockade. ΔGp: Mean conductance blockade.

In summary, we fabricated Si3N4 membranes with thicknesses of approximately 3nm across a wafer with extremely low variation in thickness. We thus believe that the poly-Si sacrificial layer process is a promising approach for fabricating ultra-thin membranes for solid-state nanopores. More detailed information on this study is available elsewhere.6 Our next step will be to discriminate the four types of nucleotides in DNA using the fabricated nanopores.

Itaru Yanagi
Hitachi Ltd.
Tokyo, Japan

Itaru Yanagi joined the Central Research Laboratory at Hitachi Ltd. in 2005. From then to 2009 he was engaged in development of semiconductor flash memory devices. Since then he has been developing biosensors. His research mainly focuses on nanopore semiconductor sensors for DNA sequencing.

1. D. Branton, D. W. Deamer, A. Marziali, H. Bayley, S. A. Benner, T. Butler, M. Di Ventra, S. Garaj, A. Hibbs, X. Huang, The potential and challenges of nanopore sequencing, Nat. Biotech. 26(10), p. 1146-1153, 2008.
2. B. M. Venkatesan, R. Bashir, Nanopore sensors for nucleic acid analysis, Nat. Nanotechnol. 6, p. 615-624, 2011.
3. R. Tewhey, V. Bansal, A. Torkamani, E. J. Topol, N. J. Schork, The importance of phase information for human genomics, Nat. Rev. Genet. 12, p. 215-223, 2011.
4. M. Wanunu, T. Dadosh, V. Ray, J. Jin, L. MacReynolds, M. Drndic, Rapid electronic detection of probe-specific microRNAs using thin nanopore sensors, Nat. Nano-technol. 5, p. 807-814, 2010.
5. A. T. Carlsen, O. K. Zahid, J. Ruzicka, E. W. Taylor, A. R. Hall, Interpreting the conductance blockades of DNA translocations through solid-state nanopores, ACS Nano 8, p. 4754-4760, 2014.
6. I. Yanagi, T. Ishida, K. Fujisaki, K. Takeda, Fabrication of 3-nm-thick Si3N4 membranes for solid-state nanopores using the poly-Si sacrificial layer process, Sci. Rep. 5, p. 14656, 2015. doi:10.1038/srep14656
Sign in to read the full article
Create a free SPIE account to get access to
premium articles and original research
Forgot your username?