An Investigation of the Performance of Comminution Vials and Ball Pestle Impact Grinders
Reprinted from American Laboratory, Oct., 1983
By Dr. Monte J. Solazzi
Centrifugally accelerated ball pestle and sample forced through a figure-eight path of travel.
IN THE PAST, powdered sample materials were comminuted in any available vials by the ball pestle impact grinding technique. The choice of vials used was limited to the familiar flat interior ended medicine-type vials with a selection of screw or snap closures. These grinding containers were not designed or necessarily suitable for providing the reproducible particle size reduction required for high degrees of spectrochemical analytical accuracies. Lack of availability of any other type of comminution vessel and insufficient understanding of the actual events occurring during the comminution process with ball pestle impact grinders, contributed to the popularity of flat interior-ended vials. This paper will discuss particle size reduction of various types of sample materials using SpectroVial® (Chemplex® Industries, Inc.) grinding vials.
Background
Chemplex SpectroMill Ball Pestle Impact Grinder
A ball pestle impact grinder is an electromechanical device designed to rapidly and energetically propels a ball pestle contained in a cylindrical vessel with a powdered sample from one end to the other. The principle of operation of a ball pestle impact grinder is based on the unique behavior of a ball pestle and its effect on a sample. The SpectroMill® ball pestle impact grinder (Figure 1) motivates a ball pestle and sample through a longitudinal figure-eight path of travel that simultaneously rotates in a 3600 pattern from one end of the grinding vessel to the other (Figure 2). However, as the ball pestle approaches one end of the grinding vessel, the grinder abruptly reverses direction opposed to the travel path of the ball pestle and sample, which increases the intensity of impact. The sample is crushed against the interior ends of the vial by the ball pestle. The cycle is repeated until a predetermined time for attaining a desired particle size has elapsed.
The degree of particle size reduction is related to comminution time, volumetric capacity of the grinding vessel, sample quantity, and type of powdered sample. To control comminution time precisely, the SpectroMill incorporates a 60-min interval timer programmable in increments of 1 min. For repetitive sample preparations, the electronic timer is set to a pre-calculated time. Depressing a reset button located within the timer dial performs re-establishing the same comminution time for subsequent sample processing. The unit automatically activates for the previously set time duration. A separate button switch controls manual operation, which overrides the automatic programmable timer.
SpectroMill-II Simultaneously Processes Two Similar or Dissimilar Samples
Flat Interior-Ended Grinding Vials
Figure 3. Traditional flat interior-ended vial with snap cap.
The flat interior-ended conventional grinding vials form nearly perpendicular angles with the interior walls (Figure 3). The spherical ball pestle cannot crush any sample particles that collect and cohere in these pockets on both ends of the vial. Similarly, the parting line established at the point of conjunction between the vial body and cap presents another area in which sample particles accumulate and avoid comminution. The parting line is usually at the edge of the open side of the vial immediately adjacent to the impact site. Under the thrust of the rapidly moving ball pestle, sample particles are forced and compacted into the crevice formed by the parting line and protected from ball pestle impact. Complicating this issue further, when a disposable plastic grinding vessel is used the closure is generally a completely different and less rigid type of plastic; the vial body is usually rigid polystyrene and the closure is a more pliable polyethylene plastic. The more pliable polyethylene plastic closure becomes momentarily distorted upon impact by the ball pestle and increases the gap of the parting line, and sample particles thus become entrapped and evade the comminution process.
SpectroVial® Comminution Vials
Figure 4. SpectroVials with concaved interior ends in both the vial bodies and closures.
The SpectroVial comminution vial was researched, engineered, and manufactured to serve one specific function: to reduce the sample preparation error for processing powdered samples in ball pestle impact grinders. The vials are cylindrical vessels with concaved interior ends in both the vial bodies and closures (Figure 4). The concaved interior ends form infinite sites to allow impacting and milling to occur, eliminate areas in which sample particles may become entrapped and avoid comminution, and promote intimate particle intermixing since the sample is continuously encouraged to motivate by the figure-eight travel path of the ball pestle. Additionally, the parting line between the vial body and closure is a considerable distance away from the ball pestle impact site to avert the accumulation and compaction of sample particles. Close manufacturing tolerances also limit the gap of the parting line to further reduce the likelihood of sample particle accumulation. In a given cycle, as the ball pestle departs from its arc of travel guided by the radial end of the vial, its rate of speed is centrifugally accelerated. Concurrently, the mechanics of the SpectroMill grinder develops a reversal of direction and imposes a sudden thrust on the ball pestle and sample to further accentuate their speed for impact and milling on the opposite end of the vial. This process continues until the sample is satisfactorily reduced in particle size.
SpectroVials (Figure 5) are manufactured in rigid clear polystyrene plastic with polystyrene friction fitting snap-on caps, and in stainless steel with screw caps fabricated of the same metal for both open ends of the vial body (Figure 6) to facilitate and ensure thorough cleansing. The polystyrene vials are disposable and the stainless steel units are, of course, reusable. Currently under investigation for use as a vial is heat-treated titanium-carbide, which is similar to tungsten carbide in hardness but not brittle and not as expensive to manufacture.
Figure 6. Metal SpectroVials
Figure 5. Disposable polystyrene SpectroVials
Experimental Statistical Analysis
The performance of SpectroVials and flat interior-ended vials was examined by comminuting sand, silicon dioxide, in a SpectroMill programmed for a fixed time duration. The ground sample material for each test aliquot was collimated through a 44 m screen, collected, weighed, and expressed as a percentage of sample quantity equal to or less than 44 m in particle size. Sand was selected as the experimental material for this application for several reasons: relative hardness and resistance to particle size reduction particularly in polystyrene vessels; initial coarse particle size of 149 m with only 5 wt% passing through a 44 m screen; and abundance of sample material for similar comparisons. Disposable polystyrene SpectroVials with a volumetric capacity of 30 ml and comparable flat interior-ended vials and two methyl methacrylate 11 mm diameter ball pestles were used for each test sample. The SpectroMill was programmed for automatic operation with a comminution time of 25 min for processing each test aliquot. After each completed grinding cycle, the comminuted samples were weighed and sieved. The collected amounts were again weighed and expressed as percentages of material passing through a 44 m screen. Ten replicate 5-g sand samples were processed in polystyrene SpectroVials and a duplicate test series was similarly prepared in polystyrene flat interior-ended polystyrene vials. A statistical analysis was performed to determine the variations in processing both groups of samples. The data are tabulated in Table 1, and Eqs. (1) and (2) were used to calculate the standard deviation, a, and coefficient of variation, v, for each test group. Table 2 shows the calculated variations for the SpectroVial and flat interior-ended vials.
where a = standard deviation
Xi = the ith individual value
n = number of observed values
x = arithmetic mean.
Examination of the data demonstrates the excellent performance and reliability of SpectroVials in statistically reproducing the weight percent quantity of test material collimated through a 44-µm screen, 19.9% 0.25, and the poor performance of flat interior-ended vials, 17.6% +/- 1.5.
Comparison Study
The experiment with silicon dioxide was extended to include evaluation of the effects on particle size by varying processing time, grinding media, and sample quantity. This investigation used two stainless steel SpectroVials of different volumetric capacities and disposable polystyrene SpectroVials. The parameters tested for each group were varied.
The first test group used a 90-ml stainless steel SpectroVial, 66 mm long x 54 mm in diameter, and contained a 15-g sample for each test. Because of the sample quantity employed, two stainless steel ball pestles of 12.7-mm diameter were used. The first sample was processed for 10 min and subsequent samples were ground in increasing 10-min increments. The wt% of ground silicon dioxide passing through a 44-µm screen was calculated (Table 3) and plotted against processing time (Figure 7). The data show excellent correlation and illustrate that within a 30-min cycle, 95.9 wt% of silicon dioxide is equal to or less than 44 µlm in particle size.
The second test group consisted of 5-g aliquots comminuted in a 35-ml stainless steel SpectroVial, 76 mm long x 33 mm in diameter, containing one 11-mm-diameter stainless steel ball pestle. It was also determined that a 5-g sample aliquot was adequate to present for spectrochemical analysis. The small size SpectroVial was commensurate with the smaller quantity of sample material processed. This group of samples was processed in increments of 5 min up to 15 min and collimated. Within a I5-min processing time, 96.2 wt% passed through a 44-µm screen. The data for this study also display excellent correlation, as tabulated in Table 4 and illustrated in Figure 8.
The last test group in this series also involved 5-g silicon dioxide sample aliquots. For this application, however, disposable polystyrene SpectroVials, 75 mm long x 33 mm in diameter, and two 11-mm diameter methyl methacrylate ball pestles were used as the grinding media. Varied processing times were selected for individual tests up to 45 min and the wt% passing through a 44-µm screen was calculated (Table 5). The weight percent of comminuted sand passing through the screen representative of the arithmetic mean, X, of 19.9 wt% was taken from the statistical analysis section (Table 1) to evaluate its position in the drawn curve relative to the other points. Figure 9 demonstrates excellent correlation, including the point inserted from Table 1 (19.9 wt%).
Other Material Investigations
The Applications Laboratory at Chemplex Industries, Inc. is frequently presented with a variety of sample materials submitted from different sources for evaluation. The following section briefly describes several experiments of various types on different materials that were difficult to process. In each example, SpectroVials were used.
Blast Furnace Slags
Three blast furnace slag samples were submitted for a suggested sample preparation procedure. A typical analysis of the major constituents was disclosed as follows: 24% SiO2, 39% CaO, 7% Fe2O3 10% MgO, 10% MnO, and 7% Al2O3). The samples were received in chunks of approximately 0.5 x 0.5 x 2 cm. Two of the slag samples, which will be identified as A and B, were occluded with steel balls ranging in size up to approximately 2 mm in diameter. The third sample, C, was relatively free of foreign metallic occlusions. By weight, samples A and B contained approximately 96% and 2% steel balls, respectively; none were detected in sample C.
The difficulty in processing these three samples was related to the unwanted presence of the occluded metallic balls. In the form received, the blast furnace slag samples were unsuitable for standard comminution procedures. The chunks were first de-agglomerated with a mallet and the steel balls were removed using a magnet. The average chip size was approximately 2 x 2 x 3 mm. To ensure complete occluded steel ball removal, each sample was subjected to a 2-min process in the SpectroMill using a stainless steel SpectroVial. The extraneous steel balls were again magnetically removed.
A 15-g aliquot of sample A, which contained the greatest quantity of occluded steel balls, was comminuted in a 66-mm long x 54-mm diameter stainless steel SpectroVial with two 12.7-mm diameter ball pestles in increments of 5 min. up to 25 min. After each grinding cycle a 1-g aliquot was removed for sieving through a 44 m screen and the quantity collimated was weighed and expressed as a weight percent.
Both factions of the removed 1-g aliquot were returned to the vial for further comminution. Table 6 shows the data for this sample preparation procedure and Figure 10 provides a graphic illustration. Inspection of the data shows an unsatisfactory distribution of points surrounding the best-drawn curve. It was suspected that the sample particles during the comminution process were not adequately intermixing for ball pestle impact. The experiment was then repeated, but this time a 0.5-g SpectroMix TM grinding/briquetting aid was added to the 15-g sample in pre-measured capsule form.
Examination of the data in Table 7 shows a significant improvement in results and excellent graphic correlation (Figure 10). Both of the curves were plotted in the same graph to provide a better view of the improvement relative to the untreated sample. To further illustrate the improvement in particle size reduction realized with SpectroMix, a comparison between untreated and SpectroMix processed blast furnace slag sample A was performed (Table 8). The data were taken directly from Tables 6 and 7, and the percentage increase of collimated material in relation to each period of processing was calculated. The use of the grinding aid also facilitated cleansing operations because of its lubricious nature and reduced the procedure to a simple dry paper towel wipe.
Sample B was used to illustrate the effect sample quantity had on particle size. The SpectroMill was programmed for a 5-min comminution cycle per test. The same stainless steel SpectroVial was again used. SpectroMix powder was added to each sample aliquot in a proportion of 3.3 wt%. Four test samples of 2.5, 5, 10, and 15 g were prepared. Table 9 shows the reduction in the weight percents of collimated materials with increasing sample quantity. Figure 11 displays excellent correlation of the plotted points and demonstrate the effect of sample quantity on particle size.
The last blast furnace slag, sample C, was similarly processed with 3.3 wt% SpectroMix and was used to determine the length of comminution time required to yield at least 95 wt% of collimated material passing through a 44-µm screen. The same stainless steel vial and ball pestles were used to keep variables to a minimum. Each sample aliquot tested was 10 g. The data presented in Table 10 show that approximately 30 min was sufficient to grind the sample to a particle size in which 95 wt% passed through the 44-µm screen. Figure 12 displays the excellent correlation between each of the plotted points.
Tin Ores
The study of tin ores was particularly interesting because the analyst had already established a viable sample preparation procedure and was exploring the possibility of incorporating the SpectroMill and SpectroVials in the scheme. Specific conditions were outlined. Fixed quantities of sample (2 g) and a grinding additive (12 g) had to be maintained. Evaluation consisted of determining the length of time required for each 14-g sample/additive mixture to furnish a particle size in which at least 95 wt% passed through a 44-µm screen. Seven tin ore specimens were submitted for study. A 66-mm Iong x 54-mm-diameter stainless steel SpectroVial with two 12.7-mm-diameter stainless steel ball pestles were employed as the grinding media. Each tin ore specimen was processed in the ball pestle impact grinder, sieved, and the collected quantity expressed as weight percent. Table 11 shows the results of this study; a processing time of 35 min was adequate to comminute the tin ores to the desired particle size.
Firebrick
A single firebrick was submitted for comminution analysis. The only requirement was not to exceed a processing time of 5 min to achieve a particle size of equal to or less than 44 m. A 76-mm-long x 33-mm-diameter stainless steel SpectroVial with two 11-mm diameter ball pestles was used as the grinding media. The sample aliquot per test was 5 g after the firebrick was pulverized. According to the data in Table 12, within a 5-min comminution cycle 96.3 wt% of the sample passed through a 44-µm screen. Figure 13 illustrates the distribution of points.
Summary and Conclusion
The effectiveness of SpectroVials in comminuting samples in a ball pestle impact grinder in preparation for spectrochemical analysis has been demonstrated. The data presented in Table 1 and Table 2 illustrates the higher degrees of precision and increased particle size reduction obtained with SpectroVials. The improvement in reproducibility is attributed to the concaved interior ends in both sides of the vial, the smooth unobstructed interior walls, and the displacement of the point of conjunction between the vial body and closure from the impact site. During comminution, individual sample particles are unable to elude ball pestle impact by virtue of the interior radial ends and the absence of pockets or crevices for sample material to collect. The resultant comminuted sample is homogeneously blended and uniformly reduced in particle size and distribution with a savings in processing time.
SpectroVials are available in different volumetric capacities in both polystyrene plastic and stainless steel to permit a comprehensive range of sample material processing. The future use of heat-treated titanium-carbide materials may extend the range of sample material processing and reduce comminution time while avoiding the brittle nature and high manufacturing costs of tungsten carbide. With the use of the SpectroMiII, the analyst can precisely control particle size reduction by varying cycle time and sample charge.
The investigations described in this paper illustrate the success of SpectroVials in reducing sample preparation error. The range of sample materials that can be processed with SpectroVials is extensive, and the examples cited demonstrate typical superior performance in the comminution process.
Note: SpectroMill®, SpectroVial®, SpectroMix® and Chemplex® are registered trademarks of Chemplex Industries, Inc.
Dr. Monte J. Solazzi is President, Chemplex Industries, Inc., 2820 SW 42nd Avenue, Palm City, Fl. 34990, USA. Tel: (772) 283-2700.