As far back as 1988, I remember papers presented on analyzing semi-solids (e.g., one by James Drennen, now Dean at Duquesne) by near-infrared spectroscopy (NIRS). Unlike liquids, where you can control the pathlength or solids (tablets), where either diffuse transmission or reflection may be used, semi-solids present unique process analysis problems. The material is likely to be less homogenous than tablets and certainly less than solutions, with density variations as well as chemical differences.
Unless contained in a fixed pathlength, the often organic matrix tends to be quite susceptible to temperature variations, affecting both concentrations and scattering properties.

One deterrent to applying PAT or mere on-line analysis can be either cost or unfamiliarity. Semi-solid dosage forms are seldom based on blockbuster molecules, seldom leading to a large net return of investment. Most of the time, branded products as well as generic products are outsourced to contract manufacturers for cost-cutting measures, which again is usually a deterrent for implementation of sophisticated PAT tools. Widespread use of PAT would be contingent upon stronger regulatory enforcements and economical use of technology.

Many semisolid manufacturers, even with the advancement of PAT technologies, tend to use empirical trial-and-error methodology to address scale-up issues. Wide usage of IR and acoustic spectroscopic methods along with implementation of innovative PAT technologies, such as integrated chemometric sensors for real-time multidimensional analysis, will rest largely on the free scientific exchange of information within the industry and more sponsored research in academia.

A semi-solid can be water-based or, more often, organic solvent-based and they come in many different forms, meaning, “one size analysis does not fit all.” The different forms include:

Hydrophobic ointments. Lipophilic ointments are usually anhydrous and absorb only small amounts of water. They contain water-insoluble hydrocarbons: hard, soft and liquid parabin, vegetable oil, animal fats, waxes, synthetic glycerides and polyalkylsiloxanes.

Water-emulsifying ointments. Can absorb large amounts of water and have a hydrophobic fatty base in which wool fat, wool alcohols, sorbitan esters, mono glycerides, or fatty alcohols render them hydrophilic.

Hydrophilic ointments. Are miscible with water and are usually mixture of liquid and solid polyethylene glycols.

Creams. Homogeneous, semi-solid preparations consisting of opaque emulsion systems, either water-in-oil (w/o) or oil-in-water (o/w).

Gels. Usually homogeneous, clear, semi-solid, consisting of a liquid phase within a 3-D matrix with cross-linkage by means of gelling agents.

Hydrophobic gels. Bases usually consist of liquid paraffin with polyethylene or fatty oils gelled with colloidal silica or aluminum or zinc soaps.

Hydrophilic gels. Bases usually consist of water, glycerol, or propylene glycol gelled with tragacanth, starch, cellulose derivatives, carboxyvinyl polymers, or Mg Al silicates.

Pastes. Homogeneous, semi-solid preparations containing high concentrations of insoluble powdered substances (> 20%) dispersed in a base.

Poultices. An ancient form of topical medication also known as a cataplasma. It is a soft mass of vegetable constituents or clay, usually heated before application.
The solvent helps determine the analysis technique used. Water almost immediately dismisses mid-range infrared and leads to either near-infrared (NIR) or Raman spectroscopy as preferred modes. Concurrently, there has been a fair amount of work being done in the life sciences, using diffuse reflection NIRS. The ability to make qualitative and quantitative measurements in severely scattering and highly absorbing matrices has become far more common than it was 30, 20, or even 10 years ago.

From medical research, we learn that the effect of particles on scattering can change when the range of particles changes from medium to very small. We published a paper in 1986¹ measuring the mean particle size of powders, based on diffusely scattered NIR light. In the study, we found that the absorbance of a material is inversely proportional to the reciprocal of the mean particle size. Also discovered, was that there is a slope change at about nine microns. This effect is seen for all samples run, inorganic and organic. Since I was in the industry at the time, I didn’t have massive amounts of time to research the “why” of this phenomenon.

Twenty years later, I was “educated” at a tech meeting,² where a medical researcher was using near-infrared spectroscopy to image breast tumors. He complained about the fact that our number of fat cells doesn’t change, when we diet; they merely shrink. As a consequence of this, the incident radiation—the exact point depends on the wavelength—goes from back-scattered to forward-scattered. The former being a larger effect, giving larger baseline shifts.

So, despite the non-specific nature of NIR spectra, all physical and chemical variations are potentially measurable. With advances in both Chemometric algorithms and computer power, complex spectra of highly absorbing materials have become almost commonplace. This is necessary, since we are dealing with a normally heterogeneous sample. Yes, tablets and capsules are, by definition heterogeneous (as a Physical Chemist, I feel obligated to point out that only true solutions—solid or liquid—may be considered homogeneous), but they are either all solid or all liquid. Ointments are a polyglot of both, even a suspension of one liquid in another, usually aided by a surfactant, is a two-phase system.

Raman spectroscopy has the advantage of not “seeing” water, giving a “cleaner” spectrum of the API in the semi-solid mixture (cream/ointment/paste). In a purely quantitative situation, this allows the analyst to choose the proper wavelength, or, more precisely, wavenumber or cm-1 shift, for quantification. If a fast-enough measurement time is used, a continuous measurement of the process is possible, assuming the ubiquitous fluorescence may be overcome. The only negative would be the non-response to particle (droplet/globule) size variations. In truth, Raman isn’t purely spectroscopy, but a mix of Rayleigh, elastic, and inelastic scattering. Polymorphic variations may be determined as can crystallinity, but particle size is not an option. Of course, in the hydrophobic mixture, moisture determinations would be very difficult. 

While human flesh is not an ointment (semi-solid), there are many similarities between the two. As a consequence, analyzing semi-solids takes more understanding of the sample than clear liquids and even complex tablets. Since many salves, ointments, and lotions are organic-based, polar components are often “solubilized” with surfactants or homogenized (made smaller, so they are suspended by Brownian motion and the viscosity of the matrix) to make a pseudo-homogenous mixture.

One analogy that comes to mind is the sizing or particles and granules by milling or sieving: proper homogenization will give proper-sized droplets, allowing them to be both distributed evenly and suspended longer. Since the profit margin for semi-solids is smaller than solid dosage forms, control (i.e., not having a bad batch) is even more important. In other words, since the cost of goods sold (COGS) to the value received ratio is too low to afford to lose a batch. Because of both the size of a typical batch and the fact that a company usually makes a large variety of semi-solid formulations, there is seldom enough time to perfect or fine-tune each one’s “design space.”

It would be simpler to use NIRS or Raman to monitor and make adjustments than attempt to spend large blocks of time and money on numerous low-profit products. In fact, since the product does not need to be dried, pressed, or coated, a design of experiment would be simple to perform. The formulator could simply make small lots with varying sized droplets and use that as a template for size measurements. Similarly, the concentration of API could be measured by a similar set of API concentrations could be made with different droplet sizes for quantitative measurements.

Using simple PLS algorithms, both droplet size and API concentrations can be followed/regulated in real time. Sound too good to be true? It’s not, really. 

References

1. “Determination of Particle Size of Pharmaceutical Raw Materials Using Near-Infrared Reflectance Spectroscopy”, Spectroscopy 1 (7), 36 (1986).
2. International Diffuse Reflectance Conference, 2006, Chambersburg, PA.

---

Emil W. Ciurczak
DoraMaxx Consulting

Emil W. Ciurczak has worked in the pharmaceutical industry since 1970 for companies that include Ciba-Geigy, Sandoz, Berlex, Merck, and Purdue Pharma, where he specialized in performing method development on most types of analytical equipment. In 1983, he introduced NIR spectroscopy to pharmaceutical applications, and is generally credited as one of the first to use process analytical technologies (PAT) in drug manufacturing and development.