CHAPTER 1
Microstereolithography
SHANGTING YOU, KATHLEEN MILLER AND SHAOCHEN CHEN
1.1 Introduction
Fabrication is a critical process in making materials into functional parts and devices. Traditional fabrication technologies such as machining and molding are commonly used in macro-scale three-dimensional (3D) fabrication. However, they are not adequate for microscale fabrication. Modern micro- and nanoscale fabrication technologies such as photolithography, soft lithography, electron beam lithography, focused ion beam lithography, dip-pen lithography, and atomic layer deposition, are often 2-dimentional (2D) in nature for thin film and surface patterning.
3D printing, as an additive free-form 3D fabrication technology, has achieved great commercial success in the past decade, because of its low cost, simplicity, and versatility. Microstereolithography is a light-assisted 3D free-form fabrication technology. This technology utilizes photosensitive materials, which can solidify upon ultraviolet (UV) or short wavelength visible light exposure. By spatially controlling the exposure dose, the desired 3D structure can be fabricated. This technology evolves in two directions. One, scanning-based microstereolithography, which provides the extremely fine resolution (sub-micron scale) but slow fabrication speed (e.g., hours). Two, projection-based microstereolithography, which provides both fine resolution (micron scale) and fast fabrication speed (e.g., seconds to minutes).
Microstereolithography is a powerful tool for biofabrication. It has successfully demonstrated its capability of fabricating a wide range of biomaterials such as hydrogels, proteins, and cell-laden materials. Both synthetic and natural materials can be used to print, each having different advantages for stability, mechanical properties, cytocompatibility, and printability. Material decisions must come from the eventual application for the printed structure, as even the choice of photoinitiator can greatly impact the print.
This chapter covers basic physical and chemical mechanisms in photopolymerization, materials, devices, and systems of microstereolithography. The photopolymerization mechanism is detailed in Section 2. Materials for microstereolithography are discussed in Section 3. Scanning-based microstereolithography, including single-photon polymerization microstereolithography, and two-photon polymerization nano-stereolithography, is detailed in Section 4. Projection-based microstereolithography, including liquid–air interface polymerization and liquid–substrate interface polymerization, is discussed in Section 5.
1.2 Photopolymerization
In this section, we will focus on photopolymerization, a technique most often used to crosslink liquid state monomers or oligomers into solid state long-chain polymers. Photopolymerization uses free radicals to initiate and crosslink strands within a monomer solution to form a solid hydrogel. When paired with microstereolithography techniques, various complex structures can be fabricated.
1.2.1 Step-growth Polymerization
Two of the main types of polymerization observed in hydrogel scaffolds are step-growth and free-radical polymerization. Although photopoylmerization methods use free radicals to polymerize structures, the kinetics of step-growth polymerization can describe some of the more unique polymerizations, and thus it is important to cover. Step-growth polymerization occurs when polymer chains grow in a stepwise fashion either by condensation reactions, in which water is removed, or when reactive end groups interact. When considering simple linear chain reactions, the mechanisms and rates of all polymerization steps can be assumed as equal. Moreover, the Carothers equation (eqn (1.1)) defines the level of completion of the step-growth polymerization
xn = 1/(1 - p) (1.1)
where xn is the average chain length, and p is the conversion rate of the monomers into polymers. This is an incredibly useful equation for predicting how long the reaction needs to proceed to obtain the correct molecular weight. As one can see, this defines an exponential relationship (Figure 1.1). As one might imagine, high molecular weights become increasingly more difficult to achieve for three reasons: (1) the frequency of reactive end groups meeting decreases; (2) the frequency of side reaction interference increases; and (3) when two or more monomer types are used in the reaction, it is difficult to ensure that the starting material concentrations are equal. This third point is an issue for users creating copolymer hydrogels, such as those consisting of monomer A and monomer B, where A-A does not react, nor B-B, but only A-B. To stop the reaction at a lower molecule weight, the user has a few options. One, the reaction can be rapidly cooled at the correct time point to slow down the polymerization rate as many step reactions have high activation energies. Two, a monofunctional material can be added to "cap" the end of the polymer, preventing it from further reactions. Lastly, if making a copolymer, a stoichiometric imbalance of the starting materials can be used. For example, if more A groups are used than B, eventually the polymer will have two A end groups on a chain and the B monomer will be completely consumed preventing further reactions with that polymer. The Carothers equation can be expanded to define this scenario, shown in eqn (1.2),
xn = (1 + r)/(1 + r - 2p) (1.2)
where xn is the average chain length, p is the conversion rate of the monomers into polymers, and r is the ratio of monomer A to monomer B (NA/NB).
1.2.2 Free-radical Polymerization
In free-radical polymerization, a free radical interacts with reactive end groups to form a polymer. The basic structure of the active group is CH2=CR1R2, as the pi-bond in the carbon–carbon double bond allows it to be rearranged when exposed to a free radical. From here forward this will be referred to as the active center. In the polymerization scheme, there are three main steps: initiation, propagation, and termination. As will be expanded on later in the chapter, the initiation step is the activation of the active center when a free radical reacts with the carbon–carbon double bond. After initiation, the active center reacts with other double bonds, propagating the chain to form a polymer and transferring the free radical to a new active center. The termination step can then occur in several ways: (1) two active centers can react; (2) one active center and one free radical can react; (3) the active center transfers to another molecule; or (4) interaction with impurities or inhibitors. Chain transfer is another important mechanism that occurs in free-radical polymerization. This occurs when an active center collides with a molecule such as a solvent, initiator or monomer, transferring the free radical to the second species.
Free-radical polymerization can theoretically go to full conversion, but the interaction of free radicals with each other must be kept in mind. Eventually, free radicals will form covalent bonds with each other, stunting the chain propagation and preventing full conversion. As a general rule, the greater the free radical concentration, the shorter the chain length. The viscosity also has an impact on the rate of the conversion, as it impacts the diffusion of polymer chains through the medium. As the conversion increases, so does the viscosity, preventing chains from interacting and terminating the reaction at the same speed. This means the propagation of chains by free radicals increases towards the end of the reaction (Figure 1.1). This auto acceleration can be prevented by stopping the reaction before the initiation and propagation steps become diffusion-mediated. Regardless of conversion, the average chain length of the completed polymer shows little variation throughout polymerization. Moreover, longer reaction times may increase polymer yield but will not increase the chain length. Increasing the temperature can decrease the molecular mass, but the best way to control the polymer chain length is to alter the initiator concentration.
1.2.3 Living Free-radical Polymerization
In "living" polymerization, the termination step is suppressed, so that the free radical is continually recycled. This results in a much lower polydispersity. Most living polymerization methods use a reversible "cap" on the active site, preventing it from continuously reacting. This slows down the reaction and suppresses termination, which results in a steady increase in chain length, defined by the following eqn (Figure 1.1):
xn = [M]o/[I]o x p (1:3)
where xn is the degree of polymerization, [M]o is the initial monomer concentration, [I]o is the initial initiator concentration, and p is the conversation rate. Studies have been performed to try to improve reaction schemes for living polymerization. Two of the most common techniques are atom transfer radical polymerization (ATRP) and reversible addition fragmentation chain transfer (RAFT) (Schemes 1.1 and 1.2).
A typical ATRP reaction consists of a dormant species, such as an alkyl halide (R-X) that can be reversibly activated by a transition metal (MtzY/L) to form an active radical (R•) and an oxidized metal complex (XMtz+1Y/L). When the free radical is not capped with the metal complex, it is free to propagate until it is capped again. ATRP methods are fairly versatile and work best with monomers such as styrenes, methacrylates, methacrylamides, and acrylonitrile. ATRP can also be used in conjunction with an initiator in a similar process called "reverse ATRP".
For RAFT reaction schemes, the "capping" species contains a dithio compound and a good leaving group within its structure. The compound reacts with the free-radical functionalized polymer chain to form a dormant chain, and the leaving group is released as another free radical in solution. At an equilibrium state, the compound can react reversibly with up to two chains, rendering them dormant. This yields a much lower polydispersity. The molecular weight of the polymer can be compared to the conversion to distinguish the two types of polymerization (Figure 1.1). Translating to hydrogel scaffolds, where the free radicals create cross-linking between polymer backbones, this means that a lower free radical concentration is needed for living polymerization than free-radical polymerization.
1.2.4 Photoinitiators
In order to create the free radicals for photopolymerization, photoinitiators are added to the prepolymer solution. The photoinitiators used in hydrogels usually generate free radicals by one of two methods: (1) photocleavage or (2) hydrogen abstraction. In photocleavage, the molecule undergoes bond cleavage at C–C, C–Cl, C–O, or C–S bonds when exposed to light, generating free radicals. In method two, the photoinitiator undergoes hydrogen abstraction from an H-donor molecule, forming a ketyl radical and donor radical. When selecting an appropriate photoinitiator, the user must consider several factors including biocompatibility, solubility in water, stability, and cytotoxicity. Photoinitiators can greatly impact the print resolution, as well as cell viability. As such, preliminary studies modulating the concentration and exposure time of the print are necessary for any new photoinitiator considered.
Photoinitiators can be characterized by the wavelength at which they most strongly absorb. Although less popular, visible light photoinitiators can be advantageous if the user plans on exposing the pre-hydrogel solution for a longer period of time. Compared with other photoinitiators that activate in the UV-range, encapsulated cells can be exposed for a longer time with less risk. However, these photoinitiators are more difficult to work with, as ambient conditions require the user to work in the dark. One popular visible light photoinitiator is Eosin-Y, which generates radicals by hydrogen abstraction. It is activated at 490–650 nm and is a common choice as a cytocompatible molecule. Unfortunately, in order to generate enough free radicals, it often needs a coinitiator (e.g., triethanolamine (TEOA)) and a comonomer (e.g., 1-vinyl-2 pyrrolidinone (NVP)) included in the prepolymer solution, making modulation of concentrations more difficult to work with than some of the common UV-initiators.
UV-activated photoinitiators can be advantageous for users who desire easier ambient conditions. However, because the pre-hydrogel solution is exposed to UV, care must be taken to limit the exposure time when working with encapsulated cells in the gel. Preliminary exposure studies with cells are recommended to ensure limited cell death. Two common cytocompatible photoinitiators are Irgacure-2959 (I-2959) and lithium arylphosphanate (LAP), both of which generate free radicals by photocleavage. I-2959 remains a popular choice among users, but has some drawbacks when compared to LAP. I-2959 is not very water soluble (<0.5 wt%), and has low molar absorption at 365 nm (ε<10 M-1 cm-1), the wavelength usually used to excite UV-activated photoinitiators. As discussed previously, the lower molar absorption greatly impacts the polymerization rate, so that large amounts of photoinitiator or long, strong exposures must be used (eqn (1.1)). Due to its low solubility, I-2959 is usually incorporated at high concentrations into solely synthetic polymer systems with non-water based solvents, or exposed for long periods of time in natural polymer systems without encapsulated cells. LAP, on the other hand, has a high absorption at 365 nm (ε ≈ 200 M-1 cm-1) and is very water soluble (at least 8.5 wt%), making it more conducive for work with natural, composite natural/synthetic polymers solutions, and encapsulated cell solutions. A comparison by Fairbanks et al. between I-2959 and LAP at 365 nm with comparable intensities and initiator concentrations found that the time to gelation was nearly an order of magnitude higher using LAP as the photoinitiator. Although commercially available, LAP can also be synthesized in-house nearly overnight, making it a reasonable option for users.
1.3 Biomaterial Choice for Microstereolithography
One of the most common materials for representing the extracellular matrix (ECM) is the hydrogel, a solid material consisting of cross-linked polymer strands that can be tuned for stiffness, adhesiveness, and cell signaling potential. Hydrogel scaffolds can be printed using natural polymers, synthetic polymers, combinations thereof, and combinations with other materials such as carbon nanotubes or nanoparticles. Natural polymers are often more cell compatible, can have cell-controlled degradability, low immune response, and are good choices for studies involving direct cell encapsulation within the gel. However, synthetic polymers offer a greater amount of control over the scaffold shape, have better batch consistency, a greater range of mechanical properties and are much more robust overtime, making them good options for seeded cell studies. By combining synthetic and natural polymers or materials in specific concentrations, scaffolds can be generated that are both cell compatible and have the appropriate physical, chemical and mechanical properties to support tissue growth in a biomimetic fashion.
