The final latex particle size in an emulsion polymerization is controlled by the short nucleation stage at the start of the reaction and by the stabilisation of the nuclei during the growth stage. Nucleation depends on the formation of radicals, a process that is very variable. This variability leads to variations in the polymerisation rate and in the size of the final latex particles.
The addition of seed particles at the start of the reaction, removes the variability in the nucleation step. The polymerization rate and particle size can be easily controlled. Seeded polymerisations also give less reactor build up, reduced pebble and give more stable latices. A reduction in customer complaints showed that customers also benefited by the change. In addition, by carefully controlling the amount of seed used, it is possible to produce bimodal latices with reduced latex viscosity.
This note describes two seeded polymerisation processes, one of which was used to develop an improved latex for use in paints and the other is used to make a bimodal PVC latex for use in PVC plastisols.
Emulsion polymerization1 involves dispersing monomer(s) in water with surfactants and a water-soluble initiator. The initiator forms radicals which initiate the polymerisation producing the polymer latex, a dispersion of polymer particles (< 1μm diameter) in water. Emulsion polymerisation is used to make polymers such as artificial rubber and PVC. Aqueous polymer lattices are also used in coating applications such as paints, varnishes, adhesives and inks.
Possibly the most important step in the polymerisation process is the nucleation or seed stage within the first few % conversion. Once stable polymer nuclei are formed and are stabilised by surfactant, they continue to grow into the final latex particles. Their rate of growth depends on the number and surface area of the nuclei, so nucleation not only affects the final latex particle size but also the rate of polymerisation. This is particularly important if the polymerisation is semi-batch and involves feeding reactants during the reaction.
Initiation occurs when the water-soluble initiator, usually a peroxide or azo compound, decomposes to form radicals, usually as a result of increasing the temperature and/or the addition of a reducing agent. Initiator radicals react with monomer molecules to form oligomer radicals of increasing molecular weight:
As the molecular weight of the oligomer radicals increases, their water solubility decreases and they precipitate, often aggregating with other insoluble oligomers to form the nuclei which grow into the final latex particles.
Initiation depends on many factors, such as the presence of dissolved oxygen, impurities and inhibitors, many of which are difficult to control. As a result the initiation step is very variable and the rate of polymerisation and final latex particle size are difficult to control.
The seeded polymerisation process overcomes these problems by providing nuclei on which the polymer particles can grow.
The seed nuclei can be any small particles that can grow to give the final latex particles. If the number or surface area of the seed particles is large enough to sweep up all of the precipitating oligomer radicals, the final latex will be monodispersed2. The number of seed particles added determines the number and (because the total weight of monomer is fixed), the size of the final latex particles.
Seed particles need not have the same composition as the final latex. They need to be small and colloidally stable in the polymerizing medium. Experiments have shown it is possible to use up to about 5%wt/wt of a seed of different composition without affecting the properties of the final latex (eg MFFT). The particle size distribution of the seed will be reflected in the size distribution of the final latex.
In this study the final polymer latex consisted of 0.2µm (200nm) diameter particles used to manufacture specialist waterbourne paints. The seed latex was a carboxylate-stabilised methacrylate copolymer latex which could be made reproducibly with a particle size of about 70nm. Because the latex was stabilized by –COO- groups, the pH of the seed latex had to be raised to pH > 6 to give good control of final latex particle size. Presumably raising the pH of the seed latex ensures the seed particles do not coagulate in the more acidic main polymerisation media.
It can be seen that the relationship between the diameter of the final latex particles and the weight of seed, of a given size, is linear and very reproducible. Thus increasing the weight of seed particles decreases latex particle size. Each seed particle gave rise to about two latex particles, suggesting some in-situe nucleation still occurred.
Compared to the original polymerization process, the seeded process gave more reproducible polymerisations with better particle size control. Build up and pebble (grits) in the reactor significantly reduced. An unexpected benefit was that it improved product quality reducing customer complaints.
As with many production changes, it was necessary to carry out a carefully designed experimental programme to show that the performance of formulated paints remained unaffected.
The viscosity of PVC Plastisols depends on the stabilizing surfactant and the PVC particle size. PVC particle size depends on the size of the spray dried and milled agglomerates and on the primary latex particle size. For a given surfactant system, increasing the PVC latex particle size or using a multimodal latex, gives plastisols with lower viscosities 3, 4.
When the number of seed particles is sufficient to sweep up all of the radical oligomers, the final latex is, as we have seen, monomodal. In seeded vinyl chloride (VCM) emulsion polymerisations containing 2x1016/kg seed PVC particles in the initial charge, a monomodal latex was obtained. As in the previous example, the size of the monomodal latex depended on the amount of seed used. But if the number of seed particles are reduced, the number of seed particles becomes insufficient to sweep up all of the radical oligomers and a second population of latex particles are formed.
The diagram below shows the particle size distribution of two bimodal PVC latices measured using a CPS Disc Centrifuge 5. The two polymerisations were identical and both used the same 580nm diameter seed latex. The only difference between the two polymerisations was that one (the red graph, PVS 168) was stopped after 30% and the other (the blue graph, PVS 169) after 72% conversion.
By knowing the weigh and the size of the latex particles it is possible to calculate their number. The table below shows how the number and size of the particles changes with increasing conversion.
It can be seen that the latex is bimodal and particle formation is essentially complete by 30% conversion. After this no new particles form. The large mode increased in size by growth and the 2nd small mode grew by both growth and by particle aggregation.
As can be seen in the graph, which shows the results from many lab scale seeded polymerisations using PVC seed latices of different sizes, the proportion of the second, smaller sized latex particles increases as the number of seed particles reduces.
By varying the amount and size of the seed latex particles charged, it is possible to control the sizes and the proportion of the two size modes. But surprisingly, although the replacement of a monomodal latex by a bimodal latex reduced plastisols viscosity, varying the size and proportion of the two modes of the bimodal latex did not appear to significantly affect plastisols viscosity 4.
For further information about seeded and other aspects of emulsion polymerisation please email us using the enquiry form on our contact page.