Mouthfeel is sensitive to microparticle size distribution so this was measured for each powder type by Malvern Mastersizer 2000 (Malvern Instruments Ltd., Malvern Worcestershire, UK) using sonication in water. Table
lists the silicon-based powder materials assessed in this study.
Silicon-based powders/material used in the study
Free-standing porous silicon membranes were fabricated by electrochemical anodisation of 6 inch diameter p
type wafers. The porous silicon flakes were rotor milled to obtain a powder with two different particle size distributions (D50 values of 4 and 23 μm). These porous silicon powders were subsequently oxidised in the air under static conditions at 600°C for 15 min. Solid silicon powders of two different size distributions (D50 values of 4 and 12 μm) were used as received from the manufacturer, as were the two porous silica powders with D50 values of 11 and 134 μm.
Gas adsorption/desorption analysis was carried out using a Micrometrics TriStar instrument (Micrometrics UK Ltd., Bedfordshire, England) with sample degassing at 70°C under vacuum for 1 hour and Brunauer-Emmett-Teller/Barrett-Joyner-Halenda (BET/BJH) analysis of the isotherms
]. Drying under vacuum was found to lower both the degas temperature and time required, compared to the flowing nitrogen treatments. The efficacy of the degassing protocol was confirmed by analysis of Davisil LC250 (Sigma-Aldrich Corporation, St. Louis, MO, USA) for which the surface area, pore volume and mean pore diameter data repeatedly agreed with the manufacturer’s specification.
Tasting of silicon powders/salts in water
The various silicon powders listed in Table
were dissolved/dispersed in drinking water (The National Forest Vending Co. Ltd., Newthorpe, Nottingham, UK, pH 6.8 at source) at different concentrations. The taste solutions of metallic salts (ultrapure grade) and silicon-based powders were prepared in drinking water according to International Standards Organization guidelines 3972:1991 using the purest salts available and the powders listed in Table
The screening tests were first performed with the purpose of defining individual threshold concentrations that can be detected. Volunteers were presented with cups coded with random numbers; each cup contained a 20 ml test solution (in varying concentrations). When a subject perceived a distinct taste sensation, he or she recorded and described the particular taste. Once the individual’s threshold was defined, in the next session the subject underwent a modified triangle test. The triangle test is a three-product test in which the task is to identify one sample that differs from the other two. Three containers of identical volume, covered with aluminium foil, were presented to volunteers and they were asked to identify the odd sample and indicate the perceived taste. The samples were coded in a random order (e.g. CCA, ABA and ABC) with one sample being the test and the other two being pure water. The major impurities of the drinking water are listed in Table
Preparation of chewing gum pellets
Si, OpSi and pSiO2-loaded chewing gum pellets were made by initially vortex mixing a 27 g batch of chewing gum base powder (Cafosa Gum Ltd., Barcelona, Spain) with 3 g of the test substance. Aliquots of 1 g were then cold pressed into pellets using a 5-mm die set with 10 kN force for 30 s.
Chew-out test protocol
The chew-out tests for metallurgical grade silicon and mesoporous silicon particles were done by a 10-person panel. The volunteers were asked to chew the pellet for 2 min. Two minute chews were followed by the collection of the saliva and rinse water for analysis of loss of silicon particles during chewing and recording the observations. The volunteers were asked to grade the samples for grittiness, taste and aftertaste.
Release of silicic acid into water and its tasting
Chew-out tests were also mimicked in vitro by mechanically grinding pellets in Tris buffer at pH 6.8/artificial saliva for 0, 2 and 10 min. The release of silicic acid in media was measured at different time points and compared with silicic acid release from pellets without grinding. The media were filtered and analysed for silicic acid content.
The taste of a silicic acid solution of 58 μg/ml in deionized water was also judged using a triangle test. Mesoporous silica (Davisil, pore diameter 15 nm, particle size 52.5 μm (D50)) was completely dissolved in Tris buffer over a 12-day period. The samples from this solution were analysed for determination of the silicic acid released in the media. From the same sample, 6 ml aliquots were presented to volunteers in sets of three, where only one sample was the test and other two were deionised water used in preparation of buffer. The volunteers underwent the triangle test again as described earlier.
Silicic acid assay
The silicic acid content was measured by spectrophotometry using molybdenum blue assay
]. It is based on the reaction of silicic acid [Si(OH)4
with molybdic acid (or ammonium molybdate) at pH 1.5 to 2 to form the yellow isomer beta silicomolybdate (SiMO12
. This molybdate complex is then reduced by sodium disulphite to silicomolybdenum blue to increase spectrophotometric sensitivity. This allows determination of silicic acid concentration in the range of 10 to 70 μgs/ml with a variance of ±10%.