Lee C C, Chan L W, Heng Paul W S
Department of Pharmacy, Faculty of Science, National University of Singapore, Singapore, Republic of Singapore.
Pharm Dev Technol. 2003;8(4):431-42. doi: 10.1081/pdt-120024696.
The process capability of a fluidized bed hammer mill was investigated with respect to four process variables, namely, rotational speeds of beater system and classifier wheel, airflow rates and length of grinding zones, as well as the particle size and flow property of the starting materials. The size distributions of all the milled lactose batches could be fitted to the Rosin Rammler distribution (RRD) function. The characteristic particle size (De) and uniform coefficient (n), which were derived from the RRD function, complemented the size at the 99th percentile of the cumulative undersize distribution (D99) to characterize the lactose batches. Lower De and D99 values indicate a finer powder while a higher n value indicates a narrower size distribution. The beater speed played a critical role. Increasing the beater speed from 12000 to 21000 rpm generally resulted in an increase in n and a decrease in D99 values due to the greater amount of milling energy supplied. The particle size and flow property of the starting material also played an important role at beater speed of 12000 rpm, where the lowest amount of milling energy was supplied. When a higher amount of milling energy was provided, the effect of particle size of the starting material was less significant. The other process variables exerted varying effects. Increasing the classifier wheel speed from 5000 to 15000 rpm decreased the De and D99 and increased the n values of the milled lactose batches, provided sufficient milling energy was supplied to the lactose particles. Changing airflow rates from 80 to 90 m3/h generally resulted in larger De and D99 values and lower n values as the higher airflow rate provided greater airflow-induced kinetic energy that facilitated the passage of lactose through the classifier wheel. However, changing the long grinding zone to a short one did not significantly affect the De, D99 and n values of the milled lactose batches produced. Small lactose particles of narrow size distribution could be obtained using the fluidized bed hammer mill upon gaining a better understanding of the milling process.
针对四个工艺变量,即搅拌系统和分级轮的转速、气流速率和研磨区长度,以及起始物料的粒度和流动性,对流化床锤磨机的工艺能力进行了研究。所有研磨乳糖批次的粒度分布都可以拟合到罗辛-拉姆勒分布(RRD)函数。从RRD函数得出的特征粒度(De)和均匀系数(n),与累积筛下分布第99百分位数处的粒度(D99)互补,以表征乳糖批次。较低的De和D99值表示粉末更细,而较高的n值表示粒度分布更窄。搅拌速度起着关键作用。将搅拌速度从12000转/分钟提高到21000转/分钟,通常会导致n值增加,D99值降低,这是因为提供了更多的研磨能量。在搅拌速度为12000转/分钟(此时提供的研磨能量最少)时,起始物料的粒度和流动性也起着重要作用。当提供更高的研磨能量时,起始物料粒度的影响就不那么显著了。其他工艺变量产生了不同的影响。如果向乳糖颗粒提供足够的研磨能量,将分级轮速度从5000转/分钟提高到15000转/分钟会降低研磨乳糖批次的De和D99值,并增加n值。将气流速率从80立方米/小时改变为90立方米/小时,通常会导致De和D99值更大,n值更低,因为更高的气流速率提供了更大的气流诱导动能,促进了乳糖通过分级轮。然而,将长研磨区改为短研磨区,对所生产的研磨乳糖批次的De、D99和n值没有显著影响。在更好地理解研磨过程后,使用流化床锤磨机可以获得粒度分布窄的小乳糖颗粒。