Non-lithographic nanofluidic channels with precisely controlled circular cross sections

Abstract

Nanofluidic channels have received growing interest due to their potential for applications in the manipulation of nanometric objects, such as DNA, proteins, viruses, exosomes, and nanoparticles. Although significant advances in nanolithography-based fabrication techniques over the past few decades have allowed us to explore novel nanofluidic transport phenomena and unique applications, the development of new technologies enabling the low-cost preparation of nanochannels with controllable and reproducible shapes and dimensions is still lacking. Thus, we herein report the application of a nanofiber printed using a near-field electrospinning method as a sacrificial mold for the preparation of polydimethylsiloxane nanochannels with circular cross sections. Control of the size and shape of these nanochannels allowed the preparation of nanochannels with channel widths ranging from 70-368 nm and height-to-width ratios of 0.19-1.00. Capillary filling tests confirmed the excellent uniformity and reproducibility of the nanochannels. These results therefore are expected to inspire novel nanofluidic studies due to the simple and low-cost nature of this fabrication process, which allows precise control of the shape and dimensions of the circular cross section.

Fig. 1. Fabrication of patterned nanochannels with circular cross sections of precisely controlled shapes and dimensions. (A) Using water-soluble NF patterns prepared by NFES as a sacrificial mold, nanochannels that are patterned over a large area can be easily formed without the requirement for cleanroom processes. During the NFES process, the spinning distance (d) and stage velocity (V_s) were varied. (B) Under dry conditions (35 RH%), the stage velocity was controlled to give NFs with circular cross sections of various diameters. Under humid conditions (50 RH%), the spinning distance was controlled to prepare NFs with different cross-sectional HWRs. (C) Schematic representation of nanochannel fabrication using the sacrificial NF mold. (D) NFs and nanochannels with mesh-type networks and parallel array patterns were examined by SEM and fluorescence imaging.

Fig. 2. Nanochannels with controlled diameters and fixed circular cross sections. (A) Schematic representations and corresponding SEM images. The nanochannel diameters were varied by manipulating the stage velocity under dry conditions (35 RH%). The corresponding NFs are also shown below the SEM images of the nanochannels. Stage velocities of 250, 300, 350, and 400 mm s⁻¹ (left to right) were employed. (B) An excellent correlation between the NF and nanochannel widths is shown (R² = 0.98). (C) Linear decrease in the nanochannel diameters upon increasing the stage velocity (R² = 0.98). (D) The circular cross section is maintained upon controlling the diameter (H and W are indicated in the SEM images of (A)).

Fig. 3. Nanochannels with controlled circular cross sections and varying HWRs. (A) Nanochannels with differently shaped cross sections were obtained by varying the spinning distance, thereby resulting in the formation of circular, D-cut, and ellipsoidal cross sections. (B) SEM images of the NFs and the nanochannel cross sections corresponding to the schematic configuration of part (A). Spinning distances of 1.2, 0.8, and 0.6 mm (top to bottom) were employed. (C) An excellent correlation between the NF and nanochannel widths is shown (R² = 0.99). (D) Variation in the NF width upon altering the spinning distance. (E) Linear relationship between the nanochannel cross section HWR and the spinning distance (H and W are indicated in the SEM images of (B)). (F) Linear relationship between the NF width and the square of the electric field, which is in turn controlled by the spinning distance.

Fig. 4. Capillary filling through the nanochannels. (A) Schematic representation of the capillary filling process. Nanochannels with controlled cross-sectional shapes and dimensions are bonded with a glass coverslip using oxygen plasma treatment, and a drop of aqueous glycerol solution is dispensed in the inlet reservoir. (B) Real-time imaging showing penetration of the aqueous glycerol solution through the nanochannels (scale bar = 10 μm). (C) Linear relationship between the position of the liquid meniscus (penetration length of the capillary flow) and the square root of time. For validation, the capillary filling test was conducted using nanochannels with four different HWRs. Average values of the experiments from five nanochannels (n = 5) are plotted.