VOC or volatile organic content has been one of the main driving forces behind coatings technological innovation and research efforts for the past 40 years since the advent of the EPA and the clean air act.1  The approaches to the problem of an ever decreasing VOC has taken two directions. First, the application of the “art” of formulation, where experience and trial and error are used to reduce the level of VOC.  At first, this approach worked but as the levels of allowed VOC dropped it found limits.  The second approach is to use “science” to define what will work and the true rules.  This second approach is the one we use at the Missouri S&T Coatings Institute.
The importance of the use of solvents in coatings has been over blown.  All solvents cost money and are only there to deliver the coating to the substrate, then, they leave.  All solvents cost to ship including water and we should look at solvent as something we would want to eliminate if we can.
The resin and the pigment, including the extender pigments which help lower the coating cost, are the two major paint components influencing the viscosity or flow of paints.  The pigment increases the viscosity as its concentration increases.  As the aspect ratio of primarily the extender increases from a sphere, ratio of 1, to platy talc with its ratio of 40, the viscosity will increase significantly at the same volume fraction.  Thus low PVC and low aspect ratio will yield the lower viscosity and aid with VOC reduction.  The most influential is the resin which is the focus of the paper.2
Resin and Solvent Interaction Issues
The resin is the major contributor to the viscosity of a coating. The choice of solvent can play a major role in how the viscosity responds to the added solvent which is demonstrated by the Mark-Houwink equation3, Equation 1.  In a good solvent, the resin usually is in a random coil configuration with an ‘a’ of 0.4-0.7.  If the polymer has charges on the chain, they repel one another and the chain becomes a rigid rod with an exponent of 2.  If the polymer chain is in a ball shape with no solvent inside (also known as hard sphere), the exponent is 0 and thus the viscosity becomes low and independent of molecular weight.4  Examples of hard sphere resins are latex, dispersions and water reducible resins including CUPs.
The viscosity of a resin in a good solvent can be controlled by adding a poorer solvent which reduces the value of exponent ‘a’ and gives viscosity which allows spray.  This technology is simply a reduction with solvent which reduces the viscosity by more than simply adding more of the good solvent since the exponent is lowered.   

n = k* Ma
The Mark Houwink equation also indicates that if we want to lower the viscosity we can also reduce the molecular weight.  However, as the molecular weight is lowered below approximately 100,000, the tensile strength drops.  Thus, although the viscosity is now sprayable at let’s say 40,000 MW, the mechanical properties of the coating are inferior and cannot be used.5  The answer to this is reaction cure “cross-link” the resin after application.  This is how 2K urethanes, urea, and epoxy coatings work by using building blocks of resins with 200-1500 molecular weight which combine many times to reach a molecular weight giving good mechanical performance.6  In addition, reactive diluents can be used which are actually monomers, which act as both a solvent and are incorporated into the resin.
WR vs CUP: a new approach
The term water reducible (WR) has been around since about the mid-1950s.  It refers to a process by which a water insoluble resin is dissolved in a high boiling, water soluble, organic solvent and base is added to form a salt.  Most of the chain is hydrophobic.  When water is added, the chains collapse into a hard sphere mainly containing only one polymer chain per particle depending upon the polymer concentration when water is added.7 
The term CUP was coined a few years ago in our research group.  It stands for a colloidal unimolecular polymer particle dispersed in water without solvent.  The process to form CUP particles is very similar to that of water reduction.  The major differences are the use of a low boiling organic solvent, lower than water, and after reduction, the solvent is stripped off to yield a VOC free resin suspended in water, Figure 1.8  This process is the true pinnacle of what we referred to as reduction in lowering the exponent of the Mark Houwink equation to zero.
The measured and calculated (from GPC data) diameters and distribution of CUP particles are virtually identical indicating a true unimolecular polymer particle.9  These particles are thermodynamically stable suspensions unlike latexes which will settle out over time.10 A study of effect of molecular weight variation on CUP formation and stability has shown that CUP particle size tracked the molecular weight except at very low molecular weight. From MW 13k up, the theoretical and experimental particle size data were in very good agreement. The CUP of 153K MW also went well and was of zero VOC.  For conventional water reducible resins, as the molecular weight goes up so does the VOC.  For example, water reducible resin with 20K MW, the VOC is about 1lb/gal and at 40K it is about 3.2 lbs/gal making it impossible to use higher molecular weight resins with the VOC restrictions we have today.11
The viscosity of CUP systems are dependent upon concentration and the amount of charge on the particle.  As the CUP concentration increases the distance between particles decreases and the charges on the particles repel each other which increases the viscosity.  The higher the charge density the higher the viscosity.  Thus, keeping the charge to the minimum needed to stabilize the particle in solution should be targeted to give the lowest viscosity.  As the particle size increases, the point at which the viscosity rises, decreases in terms of percent solids.12
As shown in Figure 2, the particles with surface water and charge can randomly pack or form a close pack configuration.  In the end, the resin must conform to the Kepler conjecture which limits the volume solids before gelation. The thickness of the bound water layer (λ) as indicated in Figure 2, is approximately 0.56 nm.  This layer is approximately the same for the three resins depicted in Figure 3 which represents a 100nm latex, a 25nm urethane dispersion and a CUP particle.
Solvent reduction can be achieved by using various methodologies and is an active field of research.   With unique modifications to the existing water-reduction technique we have developed novel aqueous polymer systems (CUPs) with zero VOC to address the solvent issues and paved the way   for future developments in resin technology. CW

References
1 K. W. McCreight, R. Stockl, C. Testa, K. S. Seo, Progress in Organic Coatings, 72, 2011, 102-108.
2 P. A. Steward, J. Hearn, M. C. Wilkinson, Advances in Colloid and Interface Science, 86, 2000, 195-267.
3 T. F. Tadros, Colloids in Paints, Volume 6, Wiley-VCH, 2010.
  4 P. C. Painter, M. M. Coleman, Fundamentals of Polymer Science, 2nd Edition, Technomic Publishing Co. Inc., 1997.
5 P. C. Hiemenz, Polymer Chemistry: The Basic Concepts, CRC Press, 1984.
6 K. Jindal, D. Bhattacharya, L. F. Francis, A. V. McCormick, L. T. Germinario, C. Williams, Progress in Organic Coatings, 67, 2010, 296-301.
7 A. R. Marrion, The Chemistry and Physics of Coatigs, 2nd Edition, Athenaeum Press Ltd., 2004.
  C. J. Riddles, W. Z hoa, H, Hu, M. Chen, M. R. Van De Mark, 8 Polymer, 55, 2014, 48-57.
9 J. K. Mistry, A. M. Natu and M. R. Van De Mark, J. Appl. Poly. Sci., 2014 DOI: 10.1002/APP.40916. 
10 M. Chen, C. J. Riddles, M. R. Van De Mark, Langmuir, 29, 2013, 14034-14043.
11 M. R. Van De Mark, A. M. Natu, S. V. Gade, M. Chen, C. Hancock, C. J. Riddles, J. Coat. Technol. Res., 11, 2013, 111-122.
12 M. Chen, C. J. Riddles, M. R. Van De Mark, Colloid Polym. Sci., 291, 2013, 2893-2901.