When I was teaching, one of the questions that used to drive me crazy was, “We learned this last year; do we still need to know it?” I assured students that science lives on “remembering” what the current ideas were based upon. Hopefully this column will help answer that question.
For some time, I have been talking about continuous manufacturing (CM) as the “next great step” in solid dosage form production. But there are other tools that make the CM process more effective. Some of the ideas were covered previously (re: the work at Rutgers to characterize raw materials), but some other techniques are either new, updated, or have merely been forgotten in the rush to perform process QbD.
One tool is an NIR imaging system by Middleton Labs that “sees” through the base of a blender (as it spins), qualifying a mixture in real time. The system produces images (hyperspectral picture), calculating the homogeneity and agglomeration in real time. They offer the sisuCHEMA as a NIR Chemical Imager, using a short wave IR camera (384 pixels x 288 spectral bands). At 400 frames/s, it is mounted on a scanner with a line illumination system. It is a tool for formulators and OOS investigators to scan a tablet, determining API and/or excipient distribution or to scan blister packs for continuity, production or clinical tablets.
Another (Middleton) tool for lab work is in fluorescence: the macroPhor combines (pushbroom) hyperspectral and fluorescence imaging. The software generates a spectrum at any pixel in the 2D image while the optics are optimized for macro-sized samples. Using an excitation of 488nm, the instrument captures fluorescence spectra from 500-800nm. The sample stage holds petri dishes, well plates or solid samples, placed on the stage, allowing API and excipients to be determined.
TeraHertz has grown from a curiosity to a nice tool in recent years. It is applied in monitoring roller-compacted ribbons (density, thickness, API distribution), following tablet coating, and a diagnostic tool for coating integrity (adherence, cracking). In-process is good, but don’t forget the scale up and diagnostic applications. With LASERs and computers, these “new” techniques are becoming workhorses. The technology is used to measure physical parameters; it may be used where NIR or Raman is now utilized in RMID.
This revolution carries over into Raman, demonstrated by Timegate instruments. Timegate’s pulsed LASER, followed by a gated detector, allows the investigator to use a shorter (visible) wavelength LASER while blocking the resultant fluorescence, which normally could swamp a Raman signal. Quite good for biopharma applications (since Raman is not bothered by water), it is also a nice tool for powder blends and tablets (and capsules).
A rapid way of gleaning information of a tablet’s interior is via a nice, new Raman-based device that can generate a 3-D picture of a solid dosage form (H2Optix). It slices layers from a tablet, scans the new surface, images it, and proceeds slicing and scanning the entire dose. This generates an internal structure of the tablet, showing the distribution of APIs and excipients. Named the “Pillerator,” it can provide automated sectioning of 5–1000 µm layers of solid dosage formulations and generate hyperspectral scans of each layer, building a computer model, or 3-D chemical map and structural analysis of a tablet or other solid analyte.
It may be used to characterize solid dosage formulations for API distribution uniformity and/or aggregation and quantify the API and/or other individual component concentrations down to 0.001% by weight. It can also estimate the particle size distribution by component, as well as perform a nearest-neighbor, long-scale correlation, and coverage percentage analysis. It is also a decent tool for prediction of dissolution profile from the solid form.
In short, what I am espousing is not returning to the “good old days,” but rather including the newer stand-alone technologies (“stand-alone” meaning not included as “in-process” monitors). I am excited about process monitors and continuous processing, but it is equally important to continue looking for new lab/pilot plant technologies that add to our knowledge of the materials that go into the process stream.
What’s old is new again.
