{"id":20368,"date":"2024-09-09T14:17:57","date_gmt":"2024-09-09T14:17:57","guid":{"rendered":"https:\/\/makingpharmaindustry.it\/?p=20368"},"modified":"2024-10-02T10:13:18","modified_gmt":"2024-10-02T10:13:18","slug":"optimising-tumble-blending-using-a-miniaturised-wireless-near-infrared-spectrometer","status":"publish","type":"post","link":"https:\/\/makingpharmaindustry.it\/en\/manufacturing\/api-excipients\/optimising-tumble-blending-using-a-miniaturised-wireless-near-infrared-spectrometer\/","title":{"rendered":"Optimising tumble blending using a miniaturised wireless near-infrared spectrometer"},"content":{"rendered":"

1. Introduction<\/span><\/b><\/p>\n

Physical characteristics of pharmaceutical blends are important in preventing powder flowing and in minimising dosage deviations in capsule filling and tableting. The content of the active pharmaceutical ingredient (API) in the final dosage form is extremely important to guarantee adequate end-product quality.<\/span><\/span><\/span><\/span><\/p>\n

Blending is a process where two or more compounds are mixed to achieve a homogenous product. To achieve that, three main powder-blending mechanisms are involved: convection, diffusion and shear, which often interfere with each other according to the type of blender used.<\/span><\/span><\/span><\/span><\/p>\n

Tumble blenders accomplish mixing by allowing free-flowing material to move within a partially filled rotating container. The capacity usually represents about 50% of the internal volume of the blender [<\/span><\/span>1.<\/span><\/span>].<\/span><\/span><\/span><\/span><\/p>\n

Powder mixing is influenced by mixing time, particle size, shape, density and type of powders as well as the design of the mixer used [<\/span><\/span>2.<\/span><\/span>]. During mixing, different particles can behave differently and the mixtures undergo mixing and demixing in successive stages during this process [3.]. The particle size distribution of the ingredients used in a powder mixture is one of the key factors that determine the homogeneity of the mixture [4.] and also affects other powder properties such as bulk density, flowability and compressibility [5.,6.].<\/span><\/span><\/span><\/span><\/p>\n

In industrial practice, blending quality is often determined by sampling with a thief probe, followed by an off-line analysis of the sampled material. Another method is near-infrared (NIR) spectroscopy that allows non-invasive and on-line measurements. By combining NIR spectroscopy and multivariate methods, critical quality characteristics of powders can be determined.<\/span><\/span><\/span><\/span><\/p>\n

This experimental study investigates the mixing behaviour of amoxicillin as a high-dosed API (96% w\/w) and talc and magnesium stearate as excipients for two different filling levels in a blender (21% and 55%). We used a miniaturised wireless NIR sensor to monitor the blending process. The sensor assures high reading accuracy through the sapphire window flanged on the lid, selected as the best dead-zone-free position. The presence of an accelerometer can ensure reproducible data acquisition during each bin rotation, providing a clearer picture of what occurred during the process and how the process parameters can impact physical characteristics of the powder during blending.<\/span><\/span><\/span><\/span><\/p>\n

In summary, our study shows one of the advantages in real-time spectroscopic monitoring of blending, minimising the number of experimental tests and eliminating the possible erroneous results observed through off-line analysis performed on sampled blends once produced. <\/span><\/span><\/span><\/span><\/p>\n

2. <\/span>Equipment and Materials<\/span><\/b><\/span><\/span><\/p>\n

2.1. Materials<\/span><\/i><\/span><\/span><\/p>\n

Formulation includes API (amoxicillin trihydrate), talc and magnesium stearate. <\/span><\/span><\/span><\/p>\n

2.2. Equipment<\/span><\/i><\/span><\/span><\/p>\n

Laboratory scale Cyclops Lab equipped with a 15-litre bin (IMA S.p.A., Active Division, Ozzano dell\u2019Emilia, Italy), pilot-scale Cyclops Mini equipped with a 100- litre bin (IMA S.p.A., Active Division, Ozzano dell\u2019Emilia, Italy) and industrial-scale Canguro 1200 equipped with a 1200-litre bin (IMA S.p.A., Active Division, Ozzano dell\u2019Emilia, Italy).<\/span><\/span><\/span><\/p>\n

2.3. In-process analysis<\/span><\/i><\/span><\/span><\/p>\n

A miniaturised near-infrared (NIR) sensor (MicroNIR<\/span>\u00ae<\/i><\/span><\/sup> <\/b>PAT-W, Viavi Solutions Inc., Scottsdale, AZ, USA) was installed as the process analytical technology (PAT) unit. The sensor was located on the upper part of the lid of the bins of the batch blender. The NIR sensor utilised a Linear Variable Filter (LVF) as a dispersive element which provides a position-dependent dispersive optical element ideally suited for use in compact instruments requiring high spectral resolution, covering a wavelength range of 950 nm to 1650 nm. Raw spectra were pre-treated using a Standard Normal Variate and the first derivative via a Savitsky\u2013Golay filter (11 pt, second polynomial order), before being analysed online using the moving block method with the F-Test. The wavelength range was selected to cover nearly the entire spectra, except for the lateral parts where the derivative pre-treatments create artificial noise. This qualitative method involved collecting independent data to compare the variances of two blocks of data at a 95% confidence interval, enabling a real-time analysis of the material\u2019s chemical composition and ensuring it remained under control throughout the entire run.<\/span><\/span><\/span><\/p>\n

2.4. Powder characterisation<\/span><\/i><\/span><\/span><\/p>\n

2.4.1. Powder Density<\/span><\/span><\/span><\/p>\n

The bulk and tapped densities are measured using Ph. Eur. Method 3. Specifically, the powder was added to a 100 mL graduated cylinder and the mass was consequently determined; the cylinder was then placed on an automatic tapping device (Erweka, Germany). The Hausner ratio (HR) was calculated by dividing the tapped density (TD) by the bulk density (BD) and the Carr\u2019s Index (CI) was calculated according to equation ( 1)<\/span><\/span><\/span><\/p>\n

\"\"<\/p>\n

2.4.2. Particle Size Distribution<\/span><\/p>\n

Particle Size Distribution (PSD) of blends is measured to detect possible changes in particle size throughout the process. Measurements are conducted by using the sieve method on a Giuliani IG\/S\/1 (Giuliani Tecnologie Srl, Torino, Italy) sieve shaker equipped with a set of sieves (63 \u00b5m, 90 \u00b5m, 125 \u00b5m, 250 \u00b5m, 355 \u00b5m, 500 \u00b5m, 710 \u00b5m and 1000 \u00b5m) on a representative sample of 100 grams of blends, by sieving for 6 minutes and weighing the separated fractions of the powder.<\/span><\/span><\/span><\/p>\n

3. Detailed case description<\/span><\/b><\/span><\/span><\/p>\n

According to the current manufacturing procedure, the blend is obtained by mixing all components in a stated order for 20 minutes at 5 rpm in a 1200-litre bin occupying 21% of total capacity. The downstream capsule filling unit operation proved to be of high quality in terms of weight uniformity and efficient in terms of manufacturing speed.<\/span><\/span><\/span><\/p>\n

The aim of the investigation is to determine the ideal blending process parameters to mix an increased batch size of 450 kg in the same bin size. Consequently, the bin filling increases from 21% up to 55% (calculated on bulk density of the starting mixture). <\/span><\/span><\/span><\/p>\n

The first evaluation is carried out on a laboratory-scale blender to maximise the results and minimise the amount of raw materials involved (API and excipients). The second step is up-scaling to a pilot bin size to prove the results obtained in the first step. This is the last verification prior to performing the engineering batch in an industrial-scale bin.<\/span><\/span><\/span><\/p>\n

3.1. Laboratory scale<\/span><\/i><\/span><\/span><\/p>\n

The main objective is to achieve the blend uniformity evaluating the ideal mixing time, mixing speed and bin filling, and measuring the physical powder properties.<\/span><\/span><\/span><\/p>\n

Blend uniformity is measured through the use of an NIR spectrometer thanks to the possibility of analysing the entire blending profile from the first heterogeneous moment to the final homogeneous one. Studying the behaviour of the mixing process and how it changes, modifying the critical process parameters (CPPs), is key to understanding the design space of this mixture.<\/span><\/span><\/span><\/p>\n

Analysing the CPPs one by one:<\/span><\/span><\/span><\/p>\n

    \n
  1. \n

    The bin-filling parameter is fixed hypothetically at 55% as the desired volume for production scale;<\/span><\/span><\/span><\/p>\n<\/li>\n

  2. \n

    The blending time is fixed at 30 minutes, which is considered long enough to fully capture the way the powder behaves and still plausible for a real blending process;<\/span><\/span><\/span><\/p>\n<\/li>\n

  3. \n

    The blending speed is the factor used to understand how blend uniformity changes inside the design space. The values selected for the blending speed are 5 rpm, 10 rpm, 15 rpm and 20 rpm.<\/span><\/span><\/span><\/p>\n<\/li>\n<\/ol>\n

    Four preliminary runs are carried out using different mixing speeds while mixing time and bin filling are kept constant.<\/span><\/span><\/span><\/p>\n

    Then, two additional runs are carried out to evaluate:<\/span><\/span><\/span><\/p>\n