Methods and Materials:

2.5.1 Preparation of film and coated urea

The preparation of the RLP films and coated urea were carried out in a one litre 100 mm round screw cap glass jar which was charged with 300 g of granular urea and preheated to 80oC in an oven. Following preheating sequential additions of 1.5 g of poly-

diisocyanate (Endurathane R100-A) and 1.5 g of polyol(composed of 10% w/w triethanol amine (TEA) in castor oil) were made. The sequence of application was repeated at five minute intervals, with the jar and contents being orbitally shaken, to allow good mixing of the two reactive components over the surfaces. The coating jar was then returned to the oven at 80oC and rotated at 30 sec intervals while the resin cured for 4 min. This coating process was repeated 5 times to produce a 5% coated product.

A number of amendments were made to the polyol mixture to increase the permeability by:

 the addition of canola and soya bean oils to increase the levels of unsaturated triglycerides and promote steric voids in the coating film

 the addition of palmitic acid to act as a foaming agent and reduce hydrophobicity and increase pore size.

58 These amendments were introduced to the castor oil /TEA mixture at 10, 15 and 20% by weight.

On completion of the coating operation, the coated urea and jar were allowed to cool, the urea transferred to a sealed plastic bag and the jar left to stand for 24 hr prior to the recovery of the film. The film was separated from the internal glass wall by carefully lifting the top edge of the film at the opening of the jar and slowly applying water between the film and glass allowing the hydraulic pressure at the film: glass interface to separate the 30-45 micron film from the glass surface. The film was then cut into sheets and air dried prior to physical testing.

2.5.2 Physical testing of polymer films for Young’s modulus and tensile strength The films were cut into 5 x 50 mm strips and conditioned at 80% RH for 24 hr prior to testing on a texture analyser (TA-XTplus, Stable Microsystems, Surrey, UK) for tensile strength and elasticity as expressed by Young‟s modulus. The measurements were carried out at a draw rate of 5 mm/min. Up to 5 repeats were performed where possible. 2.5.3 Water vapour permeability W’ of laboratory prepared polymer films

The measurement of water vapour permeability was carried out at 10, 20 and 26oC using a gravimetric method based on BS3177. In this method the thickness of the film is measured using a micrometer and the film glued to the open end of a 50 ml

polyethylene cell (45mm diameter). Following curing of the glue, a constant humidity solution of 25 ml of 50% urea solution in water was injected into the cell through a 1 mm diameter hole in the upper portion of the side wall and sealed with Bostic sealant. The cells were placed in a constant humidity (0% RH silica gel or 100% RH water) and temperature chamber for 24 hr to equilibrate. Following equilibration the weight of the cell was measured to 0.001g accuracy, and placed back in the chamber. Weighing was then repeated every 24 hr. From these results the water permeability was calculated.

59 Where w is the weight change per day in g, Vw is the specific volume of water (1 cm3/g),

lo is the polymer film thickness cm and Ø is the diameter of the film disc cm.

2.5.4 Release rate from coated urea in water

The release of urea from the coated products was determined over a 110 day period at 10 and 20ºC with a product to water ratio of 20 g to 250 ml in a sealed plastic bottle, which was shaken daily. The quantity of urea released was measured on a 0.5 ml sample of the extraction solution using a handheld refractometer and reported in % Brix. The % urea released was then calculated based on the % Brix ratio between sample and control (urea). The % Brix was found to be linear to the urea concentration with a constant of proportionality of 0.951 with R2 = 0.999.

The change in granule volume (γ) was determined at the end of the extraction period by recovering the granules from the extraction solution and measuring the granule weight and density to determine the volume of granules. This was expressed as a % increase in volume of the original granule before extraction.

2.5.5 Digital analysis of particle size and volume change distributions

The particle size distribution was measured in triplicate on 5 g samples, approximately 250 granules. The granules were carefully placed, so that they were separated, on an A4 matt black background within a calibration array of 9 x 10.2 mm diameter white plastic beads. A photo of the granules was then taken using a 4.1Mega pixel camera mounted at 45º to the surface to avoid light reflection. The image was analysed using Sigmascan Pro 5®software to measure the Feret particle diameter and maximum/minimum length of the individual particles. From these data the surface areas of the particles were estimated using an oblate approximation. The potential volume change γ is then estimated as the difference between the initial volume, calculated from the Feret radii and equivalent spherical volume and the final volume of the oblate spheroid. This analysis recognizes that only two of the three significant dimensions are recorded (length, width, height). However, the 45º of inclination of the camera effectively takes

60 an average of the two minimum dimensions. Thus, the volume calculations are

relatively accurate, and this was confirmed in later comparisons with the volume change measured by weight and density. This procedure was carried out for coated urea

granules and granules which were in equilibrium with water. 2.5.6 Film thickness distribution

The distribution of film thickness was measured on 50 randomly selected granules that had been separated from a large sample using a riffle box. The granules were

individually measured using a micrometer across the maximum and minimum diameters and weighed and then cut and placed in numbered cells in spotting tiles. Water was then added to dissolve the urea and the residual coating shells were washed, dried and

weighed. From this weight and the granule dimensions the % coating and the film thickness were determined.

2.5.7 Permeability of film with coating thickness

To investigate the effect of film thickness on water vapour and nutrient solution permeability a series of coating experiments were carried out applying 2, 3, 4, 5, 7 and 10 % coating levels of the control reactive poly-urethane resin to urea, producing mean film thickness of 0.00104, 0.00156, 0.00208, 0.0026, 0.0036 and 0.0052 cm, estimated from granule mass, surface area and coating weight.

The urea release rate from these products was then tested in water at 20oC (as described earlier) and the constant release rates and permeability of urea and thus water (Equation 2.30) determined from the release curve between 30-50%.

2.5.8 Hydraulic permeability H’

The measurement of hydraulic permeability was carried out using a pressurised

permeability apparatus which allows the determination of the hydraulic flow at varying pressures provided by a syringe and dead weights, Figure 2.2.

In this apparatus the diffusion/permeability cell is filled with 50% urea solution on the high pressure side via the pressure syringe with the air from the cell exhausted via the

61 high pressure exhaust line prior to it being clamped. On the low pressure side the cell is filled with water using the priming syringe and air from the cell exhausted via the low pressure exhaust line. With the cell primed, the high pressure syringe with platform is placed in the vertical position, the low pressure exhaust line is then primed and placed in the weighed receiver. At this point load is applied to the high pressure syringe using spotting plates (aproximatly170g each). Two minutes is allowed for the system to come to equilibrium, the receiver is then weighed and replaced under the low pressure exhaust line. The quantity of liquid passing through the membrane is then determined by

weighing the receiver 10 to 15 minutes later. The load was then further increased and flow‟s recorded until mechanical failure of the membrane or cell occurred. The pressure was then calculated based on the weight of spotting plates and corrected using a

pressure vs. load calibration curve obtained earlier using a compression tube gauge, Figure 2.3.

Figure 2.2 Membrane diffusion and permeability apparatus.

The calibration curve Figure 2.3, shows a zero offset of 14.1 kPa with a linear correlation to 120 kPa.

62

Figure 2.3 Pressure calibration curve for dead weight syringe system

The hydraulic permeability H’ and critical pressure Pcritical were then calculated from the

plot of flow (cm3 d-1) against the corrected pressure (Pa), as H’ is the slope and Pcritical is

the x-intercept.