Identify a Wide Range of Contaminants with Three Lasers

Drug products, including parenterals, must adhere to strict requirements, ensuring that they are essentially free of visible particles. Furthermore, injectable drug products must be free of visible particles that are >50 µm.

Ensuring consistency and reproducibility requires the commitment of substantial resources throughout a drug product’s formulation, development, and manufacture. Should unidentified particles be discovered, the root cause must be established and removed before production can proceed.

Inhaler, tablet, and topical cream drug products must also adhere to stringent requirements around the purity, composition, and distribution of any active particles. The distribution of excipients and active pharmaceutical ingredients (APIs) must also be analyzed to maintain consistency across lots.

The analysis of API and particle distribution can be a time-intensive stage of the drug manufacturing process. Hound offers a mix of automated microscopy, dual Raman at 532 nm and 785 nm, and Laser-Induced Breakdown Spectroscopy (LIBS) housed within a single instrument (Figure 1).

Hound counts and identifies the composition of visible and sub-visible particles with both automated and manual modes. Hound uses Raman (532 nm and 785 nm) and LIBS to identify the composition of particles, helping users track down the particle source.

Figure 1. Hound counts and identifies the composition of visible and sub-visible particles with both automated and manual modes. Hound uses Raman (532 nm and 785 nm) and LIBS to identify the composition of particles, helping users track down the particle source.

Hound facilitates the counting, sizing, and shaping of particles, as well as the ability to identify particle composition at a chemical or elemental level, assisting users in finding the particle’s source. Hound’s dual Raman lasers can identify a diverse array of organic, inorganic, and protein particles, while LIBS can identify glass, metals, and other elemental particles.

Particles can be identified, ascertaining their exact source material in a matter of minutes. Hound includes a digital camera fitted with five microscope objectives – 5x scan, 10x scan, 20x bright field/dark field, 20x LIBS, and 50x Raman – allowing the instrument to image and analyze particles with ease.

The five objectives are designed to work alongside an automated stage, and manual joystick controls, for ease of use. These capabilities mean that Hound can be used to directly identify contaminants.

Specialized filter rounds enable visible and subvisible particles to be filtered and distributed from a vial or larger volume solution. Specialized wet rounds can also be used to identity particles in suspension. Both filter rounds and wet rounds are coated with gold in order to minimize background noise while Raman spectroscopy takes place (Figure 2).

Hound uses gold coated filter rounds and wet rounds to isolate or suspend particles for counting and identification. The gold coating minimizes background interference during Raman  spectroscopy.

Figure 2. Hound uses gold-coated filter rounds and wet rounds to isolate or suspend particles for counting and identification. The gold coating minimizes background interference during Raman spectroscopy

Hound can employ gold-coated microscope slides to spread products, like topical creams, where required. The instrument’s software includes 21 CFR part 11 compliance tools throughout the data collection and analysis process. The software also includes four data collection applications to accommodate a wide range of particle analysis applications.

The Image application allows users to rapidly capture images and mosaic scans of a sample area. Particle counting features automatically provide information on particle size, number, size, and morphology. Additionally, this application enables particles that are larger than a user-defined size can be counted within a customized region of interest.

The Identify application employs Raman and LIBS to identify the elemental and chemical composition of particles based on spectral data. Spectra acquired by Hound are automatically evaluated against a built-in reference database.  

Custom databases can also be used to locate an exact composition match for the particle. The Identify application allows user to select Raman wavelength used, and change laser intensity and exposure time to further optimize any Raman spectra collected.  

Where required, users can move to use LIBS to analyze metal and other elemental particles in the same experiment. Each application includes an automated version, allowing users to load up to four samples then work on other tasks while Hound comprehensively analyzes thousands of particles within each sample.

The Hound’s automated Count, Size and Shape application employs user-defined size particle thresholding and binning parameters to count all the particles in the sample area, while simultaneously collecting morphology and size information.  

Count, Size, and Shape can be used to image the whole defined sample area, capturing high-quality images of particles depending on their shape and size before determining the size distribution of all particles in the mixture.

The automated Image-directed ID application returns the size, shape, and count of particles, identifying these with Raman and LIBS. Hound Client allows users to specify which particles to identify, based on the particle’s size, different morphology parameters, or a total number of either the largest or random particles based on their size distribution.

Every particle analyzed with Raman or LIBS can be captured in a high-quality image as part of Hound’s automated image-directed ID. Both Raman 785 nm and Raman 532 nm can be utilized to identify particles within the same experiment. With Hound, the user is able to define which Raman laser to use in particle identification.  

Users can also dictate whether or not Hound should employ its second laser to ID all particles a second time, or to focus only on particles still to be identified. After an automated analysis has been completed, verification mode allows users to re-analyze a specific particle, alter Raman wavelengths, change exposure times and/or laser intensity, or move to use LIBS should the presence of an elemental particle be suspected.  

The automated image-directed ID application can also facilitate automated LIBS analysis of several elemental particles. When using automated LIBS, particles may be adhered to the surface of a nitrocellulose adhesive round in order to avoid movement during the experiment.  

Following automated LIBS, verification mode can be used to once again re-analyze particles, or either of the Raman lasers can be employed to identify any non-metal particles.

In the studies presented here, Hound was used to manually ID a handful of visible particles discovered in a protein drug product, and automatically count and ID particles discovered in a sample. For each experiment, Raman 532 nm, Raman 785 nm, and LIBS were used as appropriate for the identification of the various types of particles.

Methods

Match criteria

Spectra from each particle were evaluated and identified using Hound’s built-in LIBS and Raman reference databases. Particles were also compared with a custom reference database that contained information on commonly used laboratory supplies. A match rank between each sample and the reference spectra was calculated by multiplying the Pearson correlation by 1000. Here, a matching rank which is higher than 700 (out of 1000) is considered to be a high-quality match.

Manual particle identification

Hound was used to analyze a protein sample that had been inoculated with three unknown, visible particles. Visible particles were extracted before being placed on a filter round. Next, these particles were analyzed using the Identify application. Raman 785 nm, Raman 532 nm, and LIBS were employed in the identification of the three particles’ composition.

Automated particle identification

A sample comprising various particles was prepared by inoculating Kimwipe fibers, a metal crimp cap particle, and plastic vial cap particles into a particle-free vial. The visible particles were then analyzed using Hound’s automated image-directed ID application.

Both visible and subvisible particles were captured using a gold filter round with 0.8 µm pores. All particles >100 µm were counted, with spectra from these particles being evaluated using a custom reference database of typical lab supplies. Raman 785 nm, Raman 532 nm, and LIBS were all employed in particle identification.

Once the automated experiment was set up, it initially analyzed particles with Raman 785 nm, before moving to Raman 532 nm. Spectra at 785 nm were acquired using a laser intensity of 38% for 60 seconds, and at 532 nm a laser intensity of 58% was used for one second.  The verification mode was employed to ID metal particles using LIBS.

Three key types of particles were identified throughout the sample. The verification mode was employed to obtain any additional spectra required from the sample during the same experiment.

Results

Manual particle identification

Three unknowns from the protein sample were manually identified by using a combination of all three lasers and the Identify application. A metallic particle was discovered in the sample, which was analyzed using LIBS. This article was identified as copper, possessing a matching rank of 939 (Figure 3A).

The fiber was also discovered in the sample and this was analyzed with Raman 785 nm before being identified as cellulose with a matching rank of 918 (Figure 3B). A further, third unknown particle was discovered in the sample, which was identified using the Raman 532 nm laser as being polystyrene, with a matching rank of 972 (Figure 3C).

Manual identification of particles inoculated into a protein sample. A: LIBS analysis of a particle in the protein sample (green) identified copper with a match rank of 939 to the reference (blue). B: Analysis of a fiber with Raman 785 nm identified cellulose (green) with a match rank of 918 to the reference (blue). C: Analysis of a particle with Raman 532 nm identified polystyrene (green) with a match rank of 972 to the reference (blue).

Figure 3. Manual identification of particles inoculated into a protein sample. A: LIBS analysis of a particle in the protein sample (green) identified copper with a matching rank of 939 to the reference (blue). B: Analysis of fiber with Raman 785 nm identified cellulose (green) with a matching rank of 918 to the reference (blue). C: Analysis of a particle with Raman 532 nm identified polystyrene (green) with a matching rank of 972 to the reference (blue).

Automated particle identification

75 large particles (>100 µm) were found after the contents of the sample vial were filtered (Figure 4). Three particle types were identified within the sample using the automated image-directed ID application.

A mosaic scan of the entire filter area showing all particles captured from the sample vial. 75 particles />100 μm were automatically counted and analyzed with Raman 785 nm, Raman 532 nm, and LIBS.

Figure 4. A mosaic scan of the entire filter area showing all particles captured from the sample vial. 75 particles >100 μm were automatically counted and analyzed with Raman 785 nm, Raman 532 nm, and LIBS.

Raman spectroscopy at 785 nm was initially employed to analyze the particles, followed by analysis via Raman 532 nm. A Raman 785 nm was used to identify cellulose fibers from a Kimwipe that were identified throughout the sample (Figure 5A).

Polypropylene (from a plastic vial cap) was also discovered within the sample, with the 532 nm Raman laser proving an ideal tool for the identification of the plastic particles (Figure 5B). LIBS revealed and identified a metal particle, which was found to be aluminum from a Wheaton crimp cap (Figure 5C).

Automated identification of particles known to come from Kimwipe fibers, plastic vial caps, or metal crimp caps. A: Raman 785 nm analysis of a fiber identified as cellulose from a Kimwipe (green) with a match rank of 927 to the reference (blue). B: Analysis of a particle with Raman 532 nm identified as polypropylene from a plastic vial cap (green) with a match rank of 965 to the reference (blue). C: LIBS analysis of a metal particles identified the particle as aluminum from a Wheaton crimp cap (green) with a match rank of 995 to the reference (blue).

Figure 5. Automated identification of particles known to come from Kimwipe fibers, plastic vial caps, or metal crimp caps. A: Raman 785 nm analysis of a fiber identified as cellulose from a Kimwipe (green) with a matching rank of 927 to the reference (blue). B: Analysis of a particle with Raman 532 nm identified as polypropylene from a plastic vial cap (green) with a matching rank of 965 to the reference (blue). C: LIBS analysis of metal particles identified the particle as aluminum from a Wheaton crimp cap (green) with a matching rank of 995 to the reference (blue).

Conclusion

Hound can employ its three lasers to successfully identify a diverse array of contaminants in a single experiment. The instrument can identify protein, elemental, organic, and inorganic particles in solution, assisting in successful identification of the root cause of any contamination.

Automated and manual particle counting and identification modes facilitate the rapid identification of a few particles with ease or the identification of thousands of particles with negligible hands-on time.

Hound utilizes three lasers, consisting of two spectroscopy types, coupled with a fully customizable reference database to successfully identify the composition of virtually any unknown particles, enabling users to quickly identify the exact source of particles in any process.

About Unchained Labs

Unchained Labs are all about helping biologics researchers break free from tools that just don’t cut it. Unleashing problem-tackling products that make a huge difference in the real science they do every day.

That’s their mantra, their promise and they own it. They live by an unconventional strategy for a start-up: they're buying commercial businesses and developed technologies, adding their magic touch to turn them into breakthrough products, investing massively in customer-facing teams and then selling those products like gangbusters.


Sponsored Content Policy: News-Medical.net publishes articles and related content that may be derived from sources where we have existing commercial relationships, provided such content adds value to the core editorial ethos of News-Medical.Net which is to educate and inform site visitors interested in medical research, science, medical devices and treatments.

Last updated: Jul 29, 2020 at 4:07 AM

Citations

Please use one of the following formats to cite this article in your essay, paper or report:

  • APA

    Unchained Labs. (2020, July 29). Identify a Wide Range of Contaminants with Three Lasers. News-Medical. Retrieved on November 28, 2020 from https://www.news-medical.net/whitepaper/20200723/Identify-a-Wide-Range-of-Contaminants-with-Three-Lasers.aspx.

  • MLA

    Unchained Labs. "Identify a Wide Range of Contaminants with Three Lasers". News-Medical. 28 November 2020. <https://www.news-medical.net/whitepaper/20200723/Identify-a-Wide-Range-of-Contaminants-with-Three-Lasers.aspx>.

  • Chicago

    Unchained Labs. "Identify a Wide Range of Contaminants with Three Lasers". News-Medical. https://www.news-medical.net/whitepaper/20200723/Identify-a-Wide-Range-of-Contaminants-with-Three-Lasers.aspx. (accessed November 28, 2020).

  • Harvard

    Unchained Labs. 2020. Identify a Wide Range of Contaminants with Three Lasers. News-Medical, viewed 28 November 2020, https://www.news-medical.net/whitepaper/20200723/Identify-a-Wide-Range-of-Contaminants-with-Three-Lasers.aspx.

Other White Papers by this Supplier