Tailoring Nanoscopic and Macroscopic Noncovalent Chemical Patterns on Layered Materials at Sub-10 nm Scales

The unprecedented properties of 2D materials such as graphene and MoS2 have been researched extensively [1,2] for a range of applications including nanoscale electronic and optoelectronic devices [3–6]. Their unique physical and electronic properties promise them as the next generation materials for electrodes and other functional units in nanostructured devices. However, successful incorporation of 2D materials into devices entails development of high resolution patterning techniques that are applicable to 2D materials. Patterning at the sub-10 nm scale is particularly of great interest as the next technology nodes require patterning of (semi)conductors and insulators at 7 nm and 5 nm scales for nanoelectronics. It will also benefit organic photovoltaic cells as phase segregation of p/n-type semiconducting polymers on 2D electrodes at length scales smaller than the typical exciton diffusion length (10 nm)

is expected to improve the charge separation efficiency [7].

Characterizing locally modulated properties of non-ovalently functionalized 2D materials requires high-resolution imaging techniques capable of extracting measurements of various physical/chemical properties. One such method is scanning probe microscopy (SPM) [18–21]. In Chapter 1, we present a brief review of SPM modalities, some of which are used to characterize interfacial properties, such as conductivity and local contact potential differences that can be modulated by amphiphilic assemblies [17, 22]. Atomic force microscopy (AFM) is one of main techniques that we use to determine topography. All imaging in this work were performed in attractive AC mode [23,24] in order to minimize disruption to the self-assembly of the amphiphiles by the scanning tip.

One challenge of using SAMs for locally modulated functionalization is that the proximity to the nonpolar interface can modify the behavior of the functionalities present on the surface in conjunction with the steric hindrance of 2D molecular assemblies. For instance, ionizable functional groups, one of the strongest local modulators of surface chemistry, undergo substantial pKa shifts (in some cases, > 5 units) at nonpolar interfaces, limiting their ability to ionize. In order to apply molecular assembly to create 2D chemical patterns, we needed to design alternative structures that can avoid such penalties against the intrinsic properties of functionalities present in the assemblies. Among amphiphiles, we observed that the chiral centers of phospholipids have the potential of elevating the terminal functional group in the head from the surface for improved accessibility. We refer to this type of assembly as a ’sitting’ phase. Chapter 2 describes sitting phase assembly of phospholipids; the projection of the terminal functionality allows it to maintain solution phase-like behavior while the dual alkyl tails provide additional stabilizing interactions with the substrates. Given the diversity of phospholipid architecture [25], the sitting phase assembly suggests the possibility of greatly diversifying the orthogonality of the chemical patterns, allowing highly precise control over surface functionalities.

While a variety of methods including drop-casting [26–28] and microcontact printing [29] have been used previously by others for noncovalent assembly of materials on the surface, they mostly address patterning scale in the sub-μm range. Here, we utilize Langmuir-Schaefer(LS) transfer, which has been historically used to transfer standing phase multilayers [30], and lying-down domains of PCDA at < 100 nm scales in the interest of molecular electronics [14, 31–33], as our sample preparation technique. LS transfer is remarkable in that the transferred molecules relinquish their pre-existing interactions in the standing phase at air-water interface to undergo ∼ 90◦ rotation and assemble into the striped phase on a substrate. This introduces the possibility of modulating local transfer rate across the substrate by manipulating local environment of the molecules. Thus, LS transfer has the potential to offer spatial control over the noncovalent chemical functionalization of the 2D substrate, essential in device applications.

In Chapter 3 and 4, We make comparative studies of various experimental factors such as surface pressure, temperature and molecular interactions that affect the efficiency of LS conversion. Considering the energetics of the transfer process, we predicted that the rate of transfer from the air-water interface to the substrate should be the highest from the regions around defects, which would be the energetically

least stable regions of the Langmuir film [34, 35]. In Langmuir films, two phases of lipid assemblies—liquid expanded (LE) and liquid condensed (LC)—often coexist at the low surface pressures (< 10 mN/m) used for sample preparation. Hence, we hypothesized that the microscale structural heterogeneity of Langmuir films could be translated into microscale patterns in the transferred film on HOPG. We compare the transfer rates between LE and LC phases and investigate the impacts of physical conditions during LS transfer such as temperature, packing density, dipping rate and contact time to conclude that local destabilization of Langmuir films leads to increased transfer efficiency. (Chapter 3)

As in the case of lipid membranes that reorganize routinely based on the structure of the constituent molecules [36–38], the structure of Langmuir films is strongly dependent on the molecular structures of the constituent molecules [39–43]. Accordingly, we expected the molecular structures/interactions to provide additional control over the LS transfer process. In Chapter 4, we compare domain morphologies and the average coverages between three single chain amphiphiles and two phospholipids, each

of which contain hydrogen bonding motifs of varying strengths. We show that by influencing the adsorption and diffusion rates, molecular architecture indeed influences LS conversion efficiency and subsequent assembly on the substrate. The presence of strong lateral interactions limits transfer and diffusion, forming vacancies in the transferred films with smaller domain sizes while weaker intermolecular interactions enabled high transfer efficiencies.