1.1
DNA-Based Construction of Molecular Photonic Devices

Glenn A. Burley

Department of Pure and Applied Chemistry, University of Strathclyde, Glasgow, UK

1.1.1 Introduction

Controlling the spatial arrangement of photonic materials reproducibly and with nanoscale precision is of fundamental importance for the development of optoelectronic devices and sensors of the future. Over the past 30 years, industry has made phenomenal progress in the fabrication of optoelectronic circuits and devices with a high level of accuracy and reproducibility. 'Top down' nanolithography has been the major driver in these developments, producing devices and circuits with resolution levels ranging from tens [1] to hundreds of nanometres [2]. Top down photolithographic approaches such as Extreme Ultraviolet Lithography (EUV) have been the predominant methods used to fabricate optoelectronic devices with sub-50 nm resolution levels. One of the major drawbacks in the further development of higher resolution circuits fabricated using EUV is the rising cost of the equipment required to produce smaller devices with sub-22 nm resolution [3]. The more recent development of Nanoimprint Lithography (NIL), for example, can replicate high-resolution patterns as small as 2.4 nm and is one of the leading contenders for the fabrication of sub-22 nm circuitry [1a], yet technological hurdles such as the defectivity and process variability of the resultant device platforms requires further development [4].

As a consequence of the increasing technological as well as economic challenges involved in fabricating devices through purely lithographic approaches, alternative methods and strategies of fabrication are now being investigated from both a fundamental as well as an applied perspective [5]. Building circuits and devices from functional molecular building blocks, that is, a 'bottom up approach', is a particularly attractive method for achieving molecular-scale precision [6]. There is increasing interest in using supramolecular assembly principles to form functional optoelectronic devices and sensors for device applications [7], yet despite a number of seminal advances in this area [8], a significant challenge still remains, that of fabricating precisely defined and error-free nanomaterials over micron-scale surface areas with complete 3D control and sub-nanometre resolution in a reproducible fashion de novo [9]. In contrast, Nature is astute at preparing micron-scale, self-assembled nanostructures via the use of a template-driven process to direct both the formation and the control of the growth of the overall nanostructure [10]. For example, the protein ferritin can be used as a template for the controlled biomineralisation of nanostructures [11]. Peptides can also be programmed to assemble in nanostructures and even act as templates for the assembly of non-natural functional materials; however, the ability to form bespoke functional materials is still restricted by our limited understanding of the rules that govern their self-assembly [10].

Of the biomacromolecules available in Nature, DNA molecules and their structural analogues have emerged as excellent templates to guide the synthesis [12] as well as the assembly of functional nanomaterials from the 'bottom up' (Figure 1.1.1a) [13]. By exploiting the predictable base-pairing rules of DNA and the high density of information embedded in its structure, DNA-programmed self-assembly can form sophisticated multi-dimensional assemblies ranging from 3D crystals [14], micron-scale 2D [15] and 3D [15b, 16] DNA nanostructures, as well as dynamic nanostructures [17], which can be reconfigured to release a therapeutic cargo in response to molecular cues [18].

Figure 1.1.1 (a) Watson-Crick base-pairing is used in Nature to store genetic information and in DNA nanotechnology to direct the assembly of sophisticated multi-dimensional nanostructures. DNA analogues such as Peptide Nucleic Acids (PNA) have also been used to direct the assembly of DNA nanostructures. (b) Schematic representation of DNA origami. A single-stranded DNA template is weaved in two- and three-dimensional DNA nanostructures using a variety of oligodeoxyribonucleotide (ODN) staple strands. (c) Triplex Forming Oligonucleotides (TFOs) offer an alternative directing modality through the formation of triplex structures

The principal aim of this chapter is to highlight the recent developments in the use of DNA-programmed self-assembly to guide the construction of discrete photonic nanostructures. The advantages and disadvantages of using DNA-programmed self-assembly to construct arrays of organic fluorophores and proteins will be presented. The second half of the chapter will review efforts focusing on different modes of DNA-programmed self-assembly to fabricate optoelectronic circuits and light-harvesting complexes. For specific applications of DNA-directed assembly for the construction of supramolecular photosynthetic mimics, the reader is directed to a recent review by Albinsson, Hannestad and Börjesson [19]. DNA-programmed assembly of metallic and semiconductor nanoparticles is another rapidly expanding area of DNA nanotechnology. This has been the subject of recent reports and will not be discussed herein [13a, 13c, 20].

1.1.2 Using DNA as a template to construct discrete optoelectronic nanostructures

DNA is a unique self-assembling molecular system. This uniqueness arises from the inherent programmability of Watson-Crick base-pairing of Adenine (A) hydrogen-bonding with Thymine (T) and Guanine (G) hydrogen-bonding with Cytosine (C) [13b]. Both the programmability and flexibility of these pairing rules can be used to form a variety of structures ranging from simple duplexes through to more complex four-stranded Holliday junctions. Further enhancement of the stiffness of DNA nanostructures is also possible as a consequence of the development of double and triple cross-over motifs [13b]. The high level of programmability of DNA is also underpinned by the availability of pre-designed sequences - both short and long. This is a key aspect of DNA nanotechnology and sets it apart from other self-assembling biomolecules as both short and long DNA sequences can be prepared, and amplified to produce suitable amounts of the template for fundamental investigations. For example, solid-phase synthesis can produce modified oligodeoxyribonucleotides (ODNs) up to ~120 nucleotides in length [21], whereas longer DNA sequences of up to 20 kilobases in length can be prepared using the Polymerase Chain Reaction (PCR) [22].

Taken collectively, both the availability of material, the predictability of self-assembly rules and the more recent advent of computer software to facilitate the design of DNA nanostructures has spurred on the construction of sophisticated two- and three-dimensional nanostructures. One of the most successful exemplars of this has been 'DNA origami', which weaves a long single-stranded DNA template with the help of a series of shorter ODN strands (Figure 1.1.1b) [15a]. Since modified ODNs with a precise functionalisation pattern can be prepared by solid-phase synthesis, the insertion of non-natural functionality at precise locations in an origami-design DNA-programmed array can be realised [21, 23], and has been used with great effect to template a vast array of functional materials along a DNA nanostructure [21, 24].

Traditional strategies to construct DNA nanostructures have focused on utilizing Watson-Crick base pairing between two complementary DNA strands. A less investigated strategy to control the addressability of optical functionality is to exploit the topological features of higher order DNA structures (Figure 1.1.2a). With its 2 nm diameter, repetitive helicity of 3.4 nm, arrangement of base-pair 'bits' of information every 0.34 nm and its widespread occurrence in DNA nanostructures, B-type double-stranded DNA (dsDNA) offers an auxiliary mode to address optoelectronic materials. For example, the large surface area and solvent accessible major groove is the primary site for DNA-binding domains found in Transcription Factors [25]. Minor-groove binding small molecules such as Hoechst 33258 (1) and DNA-binding polyamides (PAs, 2, Figure 1.1.2b) [26] offer an alternative mode of duplex DNA binding in the deep, hydrophobic minor groove [27]. Tethering functionality to specific sites on these molecules can therefore be used to direct optoelectronic materials to a specific dsDNA sequences within nanostructures.

Figure 1.1.2 (a) Structure of a B-type DNA duplex and the location of ligand binding. (b) Representative subset of DNA minor-groove binder (e.g. 1 and 2) and DNA intercalators (e.g. 3 and 4)

Nucleic acid analogues such as Peptide Nucleic Acids (PNA) and Triplex Forming Oligonucleotides (TFOs) are another family of molecules that can bind to dsDNA in a sequence-selective fashion. PNA, for example, has a number of binding modes, ranging from strand invasion of dsDNA through to triplex formation (Figure 1.1.1a) [28]....