$$\rightleftharpoonup{xx}$$
$$\longleftharp{xx}$$,
$$\longrightharp{xx}$$,
Nanofluidics is an emerging research area of µTAS (Micro Total Analysis Systems) to study biological processes or transport phenomena of ions and molecules at the length scale of 101- 102 nm. With the advent of the nanofluidic tools such as nanochannels, transport processes of molecules and ions can be monitored with unprecedented precision and manipulated, if needed, by exploiting features that are available only at this length scale for separation and detection.1,2 One of these characteristic nanoscale features is a high ratio of surface to bulk charge (or Dukhin number) in nanochannels that can cause a charge imbalance and initiate ion concentration polarization (ICP) between the nanochannel and microchannel.3
A common device platform for the study of nanofluidic phenomena consists of a two-microchannel system connected by an array of nanochannels as a junction.4-6 The material of choice for building such a nanofluidic device is the silicon because of its high stiffness that prevents the channel from collapsing during bonding processes.7 However, silicon device fabrication requires expensive masks and substantial amount of processing in the cleanroom facility.8-10 Due to the convenience of device fabrication through molding and plasma bonding, polydimethylsiloxane (PDMS) has widely been accepted as a building material for microfluidics and it would be an ideal material for nanofluidics as well. However, its low Young's modulus around 360-870 KPa, makes the PDMS channel easily collapsible during plasma bonding. The minimum aspect ratio of the nanochannel (width to depth) has to be less than 10:1 which means that the fabrication of PDMS devices via standard photolithography will become extremely challenging if the nanochannel depth has to be below 100 nm, requiring a channel width less than the current limit of photolithography at around 1 µm. To overcome this limitation, there have been attempts to create nanochannels in PDMS using non-lithographical methods such as stretching to initiate cracks with mean depth of 78 nm11 or to form wrinkles after plasma treatment.12 Collapsing a PDMS channel with mechanical pressure allowed a nanochannel height as low as 60 nm.13
Even though these highly inventive non-lithographic methods allowed building nanochannels below 100 nm in depth, the dimensional controllability of the nanochannel fabrication still poses an obstacle to a wide acceptance of PDMS as a building material for nanofluidic devices. Another critical problem of the nanochannels, whether in silicon or PDMS, is the surface functionalization in case there is a need to alter the surface charge on the channel wall for the manipulation of ions or molecules. After device assembly through bonding, the nanochannels are extremely difficult to reach for surface functionalization due to the diffusion-limited transport. To create a nanoscale channel with high dimensional fidelity and facile surface functionalization, the self-assembly method of colloidal particles induced by evaporation14-16 in microfluidic devices can be one of the promising approaches. Besides the controllability of pore size and surface property, there is even a possibility to tune the size of the pore in-situ when using colloidal particles coated with polyelectrolytes by controlling temperature,17 pH,18,19 and ionic strength.18 Because of these advantages, the self-assembly method of colloidal particles has already found applications for electrochromatography,20 biosensors,21 protein concentration22 and separation of proteins and DNA in microfluidics.14,23 In this study, we deployed this self-assembly method to build an electrokinetic preconcentration device in PDMS that requires a nanofluidic junction between two microchannels.24 The fundamental mechanism behind the electrokinetic concentration is based on ion concentration polarization (ICP).25 A detailed description of fabrication and assembly steps is included in the following protocol.