chemetrix Archives - Chemetrix https://chemetrix.co.za/tag/chemetrix/ Powerful Partnership, Inspiring Insights Thu, 30 Jul 2026 14:11:53 +0000 en-US hourly 1 https://wordpress.org/?v=6.3.10 Up the Stack: Why the Real Frontier of AI Lives Between Your Laboratory Instruments https://chemetrix.co.za/up-the-stack-why-the-real-frontier-of-ai-lives-between-your-laboratory-instruments/ https://chemetrix.co.za/up-the-stack-why-the-real-frontier-of-ai-lives-between-your-laboratory-instruments/#respond Mon, 17 Aug 2026 09:10:53 +0000 https://chemetrix.co.za/?p=7546 Continue reading "Up the Stack: Why the Real Frontier of AI Lives Between Your Laboratory Instruments"]]> A single high-resolution instrument run can generate several gigabytes of data in minutes. That sounds like progress, and it is, until that data sits isolated in a proprietary format that nothing else in the lab can read without manual intervention. This is the quiet ceiling that modern laboratories are hitting: the first wave of “smart” instrument features, automated peak integration, predictive maintenance alerts, solved real problems inside individual instruments, but they never solved the problem between them.

For lab directors, IT architects, and operations managers, this matters because the next phase of laboratory digital transformation isn’t about smarter instruments. It’s about whether the software stack above those instruments can finally talk to all of them at once.

Why do “smart” lab features stop solving problems at the instrument level?

Automated peak integration and predictive maintenance alerts were a genuine step forward; they reduced manual review and caught failures before they became downtime. But Chemetrix’s consultants have seen the same pattern repeat across regional labs: once those features are deployed, the bottleneck doesn’t disappear. It moves. The constraint shifts from “is this instrument performing well” to “can this instrument’s data talk to the next system in the workflow.”

A chromatography system might integrate its own peaks flawlessly and still hand off a result that a LIMS, an ELN, or a second instrument can’t parse without a manual export-and-reformat step. The smart feature did its job. The workflow around it didn’t get any smarter.

Why does fragmented lab data matter beyond the bench?

The instruments driving modern science, high-resolution mass spectrometers, multi-detector chromatography systems, generate data at a scale that simply didn’t exist a decade ago. A single run can produce several gigabytes of fragmented output, structured around a specific vendor’s proprietary format rather than the workflow the data is meant to serve.

This isn’t just an IT inconvenience. Every manual reformatting step is a point where data can be transcribed incorrectly, where context gets lost, and where audit trails become harder to defend. In regulated GxP environments, this kind of fragmentation isn’t a productivity issue, it’s a compliance exposure waiting to surface during an audit.

How does declarative orchestration differ from standard lab integration?

Most current lab integration is imperative: an IT team manually maps Instrument A’s output format to Instrument B’s expected input, instrument by instrument, vendor by vendor. It works, but it’s rigid. Every new instrument, every software update, every vendor change requires re-mapping the connection by hand.

Declarative orchestration works differently. An analyst sets a scientific goal, “run this sample through this workflow”, and the software layer coordinates the instrument fleet to execute it, regardless of which vendor built which piece of hardware. This shift is only possible when the data underneath is described using open, vendor-neutral standards rather than proprietary formats that lock data inside a single system.

This is where frameworks from the Allotrope Foundation, the Pistoia Alliance, and SiLA become directly relevant. The Allotrope Foundation promotes industry standards that enhance data interoperability, aligning laboratory data practices with FAIR and ALCOA+ principles to support data integrity and compliance. SiLA promotes AnIML, a basic-level data format that counters the problem of lab instruments speaking different languages, aiming to align communication between manufacturers using a shared XML-based data format. Together, these standards are what let agentic AI move from a chatbot overlay sitting on top of a single system to genuine cross-instrument orchestration

Closing the data gap for pharmaceutical and GxP labs

For regulated pharmaceutical environments, the stakes around data fragmentation are highest. A multi-step QC workflow that spans HPLC, mass spectrometry, and a LIMS needs every handoff to be defensible during an audit, not just functional day to day.

The Pistoia Alliance’s MethodDB project, developed in collaboration with the Allotrope Foundation and partners including Agilent, Merck, and Bristol-Myers Squibb, uses the Allotrope Data Format to standardise how analytical method descriptions are recorded, increasing data integrity and scientific reproducibility. This kind of standardisation is precisely what reduces the change-management burden GxP labs face when adopting any new automation layer, the data structure underneath stays consistent even as the tools on top evolve.

Chemetrix application specialists work directly with regulated labs across the region to assess where data handoffs currently rely on manual reformatting, and to map a practical path toward standards-based interoperability without disrupting validated workflows already in place.

 

Improving cross-site asset utilisation for multi-lab operations

For organisations running multiple lab sites, instrument-level smart features solve a single-site problem while leaving a bigger one untouched: knowing which instruments across the network are actually available, and routing samples to them efficiently.

This is the orchestration layer gap that the broader industry has started naming directly. Coordination “still happens in calendars and email threads” in many R&D organisations, even where individual instruments are highly automated. Declarative orchestration, built on open data standards, is what allows a scheduling system to see instrument availability across sites and route work accordingly, rather than relying on a person manually checking which lab has capacity.

For multi-site operations evaluating this shift, Chemetrix can walk through a practical asset-utilisation audit, identifying where standards-based data exchange would unlock cross-site scheduling that today depends on manual coordination.

Should labs treat agentic AI as a cure-all for workflow problems?

No, and treating it that way is precisely the steep change-management trap that catches labs out. Agentic AI is only as effective as the data structure underneath it.

Agentic AI systems capable of autonomous planning, tool use, and multi-step reasoning represent a genuinely transformative shift, but the implications come with real governance challenges: how decisions are validated, how audit trails are maintained, and what infrastructure is needed to deploy agents safely at scale. A chatbot layered on top of fragmented, vendor-locked data doesn’t solve the underlying problem, it just adds a conversational interface to the same broken handoffs.

This is the cultural shift Chemetrix advocates for: building the data foundation first, with open standards and genuine interoperability, rather than reaching for an AI overlay as a shortcut around the harder infrastructure work. Labs that get this sequence right end up with systems that are auditable, defensible, and genuinely faster. Labs that skip ahead end up with an impressive demo and the same fragmented data problem underneath.

What should lab directors and IT architects do next?

The shift from instrument-centric benches to integrated, workflow-first ecosystems isn’t optional for labs that want to capture real efficiency gains, but it does require sequencing the work correctly: assess where data handoffs are currently manual, prioritise standards-based interoperability over point-solution AI overlays, and treat regulated change-management requirements as part of the plan from the outset, not an afterthought.

For Lab Directors: Save this guide and start mapping where your current workflow relies on manual data reformatting between instruments.
For IT Architects: Talk to a Chemetrix specialist about evaluating your lab’s readiness for open-standards-based data exchange.
For Operations Managers: Request an asset-utilisation audit to identify where declarative orchestration could improve cross-site scheduling and turnaround time.

TL;DR

First-generation “smart” lab features, automated peak integration, predictive maintenance alerts, solve problems inside a single instrument, but they don’t solve the data gaps between instruments. The real bottleneck is fragmented, vendor-locked data. Open standards like Allotrope, SiLA and the Pistoia Alliance’s frameworks are what let labs move from imperative integration (manual, rigid data mapping) to declarative orchestration, where an analyst sets a scientific goal and software coordinates the fleet, improving turnaround, compliance,
and asset utilisation.

]]> https://chemetrix.co.za/up-the-stack-why-the-real-frontier-of-ai-lives-between-your-laboratory-instruments/feed/ 0 How to Get Reliable FTIR Identification Every Time with the Agilent Cary 630 https://chemetrix.co.za/how-to-get-reliable-ftir-identification-every-time-with-the-agilent-cary-630/ https://chemetrix.co.za/how-to-get-reliable-ftir-identification-every-time-with-the-agilent-cary-630/#respond Mon, 01 Jun 2026 08:35:35 +0000 https://chemetrix.co.za/?p=7478 Continue reading "How to Get Reliable FTIR Identification Every Time with the Agilent Cary 630"]]> Your raw material just failed its FTIR identity check. The match score came back at 87%, the batch is on hold and three people are standing around a spectrometer debating whether to reject a container of perfectly good excipient.

Here is the hard truth: in most cases, the material is fine. The instrument is fine. The problem is a misunderstanding of what FTIR matching actually measures, and it is costing pharmaceutical QC labs hours of unnecessary retesting, false rejections and eroded confidence in a technology that is genuinely excellent at its job.

This article unpacks the three real causes of low match scores, what FTIR can and cannot tell you, and how to build a workflow that turns your spectrometer into the trusted identification workhorse it was designed to be.

 

Why does my FTIR keep giving me low match scores?

Walk into most pharmaceutical QC labs and you will see some version of the same frustration. A technician places a powder loosely on the diamond crystal, presses the plunger halfway down, gets a poor match score and logs it as a suspected material failure. After years of working alongside QC teams across hundreds of laboratories, the Chemetrix team can tell you that this scenario almost always comes down to one of three things.

Poor ATR crystal contact: Infrared light only penetrates a few micrometres into the sample surface. Air gaps or inconsistent pressure mean the evanescent wave never reaches the material properly, producing a weak and noisy spectrum that will not match any reference cleanly.

Generic reference libraries: Most labs rely on commercial libraries built from a single reference sample under ideal conditions. When your actual material arrives with slight batch variation, a different particle size or minor additives, it will not match that idealised reference, even if it is exactly the right material.

Misplaced expectations: A match score measures similarity under the conditions of measurement. It is not an absolute verdict on material identity. Even a correct material from the correct supplier can return a score well below 100% if technique or library quality is off.

The good news is that all three of these problems are fixable.

What is FTIR actually measuring?

FTIR spectroscopy works by focussing on infrared light at a sample and measuring which wavelengths the material absorbs. Every compound has a unique pattern of absorption peaks, its molecular fingerprint. The ATR (attenuated total reflectance) method makes this fast and practical by bouncing the infrared beam off a diamond crystal in contact with the sample, requiring no preparation beyond placing the material on the sensor.

Library matching then compares your sample’s fingerprint to a reference spectrum using a mathematical correlation algorithm. The result is a match score, not a binary yes or no,
but a measure of how similar the two spectra are under the conditions of that measurement. Importantly, standard library search methods are designed to detect differences in the 5 to 10% range between a sample and its reference, which means some variation is expected
and entirely normal.

This is a critical distinction. Natural spectral variation occurs between batches of the same material due to:

  • Moisture content and surface effects
  • Particle size differences between suppliers or manufacturing runs
  • Trace additives or stabilisers that are entirely acceptable within specification

None of these make the material wrong. But all of them will affect your match score if your reference library was not built to account for them.

Here is the insight that changes everything:

FTIR is an identification tool, not a purity tool. It will tell you what something is. It will not tell you how pure it is or quantify trace impurities. That is the job of HPLC. Stop asking your FTIR to do HPLC’s work and start letting it be the world-class ID workhorse it was built to be.

How do you improve ATR sample preparation?

Consistent results require consistent technique.
These are the four rules that make the biggest difference:

Powders: Spread the sample evenly across the full crystal surface before pressing. Apply firm and consistent downward pressure using the sample press. If the signal is weak, press harder and recollect the background.

Hard solids: Press a flat face directly onto the crystal or grind a small portion to a fine powder first. A smooth surface fragment beats a large irregular chunk every time.

Pastes and semi-solids: Apply a thin even layer onto the ATR surface with a spatula. These conform naturally to the crystal and are the easiest sample type to handle well.

Crystal hygiene: Clean the crystal between every sample with an appropriate solvent and a lint-free cloth. A contaminated crystal from the previous sample is an invisible source of error that will affect match scores across multiple runs.

Standardising these steps into a written SOP, what Chemetrix calls the “SOP of the Squeeze”, is the single fastest way to improve reproducibility across your team.

How can you build a pharmaceutical FTIR library that actually works?

The solution to false rejections in pharmaceutical raw material identification is not to lower your acceptance thresholds. It is to build a smarter library. A site-specific library collects reference spectra from your own approved batches of each material, across multiple suppliers and multiple delivery lots, so that the natural variation of your actual materials is built into the reference from the start. When a new batch arrives with slightly different moisture content or particle size, your library recognises it as the correct material because it has seen that variation before.

The instrument that makes this possible: the Agilent Cary 630 FTIR

The Cary 630 is purpose-built for pharmaceutical QC applications. Its instrument configuration delivers energy throughput up to 30% greater than comparable routine FTIR systems, which means faster data collection, lower noise and more reliable spectra even when sample variation is present. Its diamond ATR crystal is impervious to abrasion, requires very small amounts of sample and is easy to clean between samples. Modular sampling accessories click in and out in seconds with no alignment required, and at just 20 x 20 cm on the bench and 3.8 kg, it lives where the work happens rather than in a centralised lab.

 

 

Most importantly, the Cary 630 runs Agilent MicroLab software, a method-driven and pictorial interface that makes building site-specific libraries an intuitive task rather than a data science project. MicroLab supports advanced classification methods including PLS-DA (partial least squares discriminant analysis). In published testing using the Cary 630 with MicroLab, a PLS-DA method achieved 100% correct classification of pharmaceutical samples and detected contaminants at levels as low as 0.5%. Results are displayed as colour-coded pass/fail indicators and automatically logged in 21 CFR Part 11 compliant audit trails.

This is what Chemetrix calls the Golden Reference Library: a site-specific and validated collection of spectra that reflects your materials, your suppliers and your real-world conditions. Not a generic database. Yours.

 

Practical resources:

What does a validated FTIR library mean for regulatory compliance?

For pharmaceutical manufacturers operating in a regulated environment, a site-specific library is not just a practical improvement. It is a compliance asset. The Cary 630 meets or exceeds the performance specifications of the US, European, Japanese, Chinese, Indian and International Pharmacopoeia. Its automated IQ/OQ qualification software generates documented and auditable performance reports covering wavenumber accuracy, spectral resolution and signal-to-noise ratio. MicroLab OQ handles qualification for the US, European and Indian Pharmacopoeia directly, while MicroLab PC can be configured for additional requirements including the Japanese Pharmacopoeia. The optional MicroLab Pharma software package adds full 21 CFR Part 11 and EU Annex 11 support including electronic signatures and data integrity controls.

 

But instrument qualification is only one part of the compliance picture. A validated reference library, with documented collection conditions, approved batch traceability and defined acceptance criteria, is what gives your QA team the scientific and regulatory confidence to act on FTIR results without
second-guessing every match score.

Chemetrix supports clients through the full validation process, from initial library design through to IQ/OQ documentation and method verification against pharmacopoeial requirements.

Practical resources:

📌 Contact Chemetrix to discuss validation support for your FTIR workflow.

There is a better way to run your QC lab

Too many pharmaceutical QC labs have quietly accepted that FTIR is unreliable. They have added extra retesting steps. They have lowered thresholds without understanding why. They have learned to work around the instrument rather than with it. That is not good science. And it is not a sustainable way to run a quality control operation. Chemetrix believes that every lab, regardless of its size, its budget or its current frustrations, deserves instrumentation that works, workflows that are documented and a team that understands what their results actually mean.

The Cary 630 is not just sold and installed. Chemetrix partners with your team through three practical engagement points:

  • Workflow audit: Identifying exactly why your current match scores are low and where the process is breaking down
  • Method development Assistance: Building and documenting the SOP of the Squeeze that standardises ATR technique across your team
  • Validation support: Ensuring your custom library meets the compliance requirements of your QC environment

When your library is right and your technique is standardised, FTIR stops being a source of anxiety. It becomes what it was always supposed to be: fast, reliable and auditable identification, every time.

 

Low FTIR match scores are almost never an instrument problem.
They are a library problem or a technique problem, and both are fixable.

The three things to take away from this article:

    • A match score measures similarity under the conditions of measurement. Natural batch variation, particle size differences and trace additives will affect it even when the
      material is correct.
    • ATR sample preparation technique has a bigger impact on match scores than most labs realise. Standardising your approach with a documented SOP is the fastest win available.
    • A site-specific library built from your own approved batches will outperform any generic commercial database for your materials, in your lab and from your suppliers.

The Agilent Cary 630 FTIR makes the instrument side of this straightforward. Chemetrix makes the rest straightforward too.

Ready to stop second-guessing your FTIR results?

📩 Contact the Chemetrix team to book a workflow audit, arrange a Cary 630 demonstration or discuss building your Golden Reference Library: chemetrix.co.za

]]> https://chemetrix.co.za/how-to-get-reliable-ftir-identification-every-time-with-the-agilent-cary-630/feed/ 0 Powerful Partnerships, Inspiring Insights. https://chemetrix.co.za/powerful-partnerships-inspiring-insights/ https://chemetrix.co.za/powerful-partnerships-inspiring-insights/#respond Tue, 27 May 2025 06:56:15 +0000 https://chemetrix.co.za/?p=6942 Continue reading "Powerful Partnerships, Inspiring Insights."]]> For over 45 years, Chemetrix has been a trusted partner to laboratories across Southern and East Africa, providing world-class analytical instrumentation and unwavering support.

As the sole authorised distributor of Agilent Technologies’ chromatography, mass spectrometry, and spectroscopy products in the region, we deliver cutting-edge solutions tailored to your laboratory’s unique needs.

Our team of highly qualified professionals—experts in the natural, life, and applied sciences—offer comprehensive support, from technical sales to application insights. We collaborate closely with our clients, ensuring that each solution aligns with their specific objectives.

Beyond our commitment to scientific excellence, Chemetrix is dedicated to social responsibility. Since 2017, we’ve partnered with eDeaf to provide skills development for deaf youth and actively participate in the Yes4Youth initiative, addressing youth unemployment in South Africa.

Whether you’re scaling up operations, navigating regulatory complexities, or seeking to enhance efficiency, Chemetrix stands ready to support your journey.

Let’s advance science together.
Connect with our team to discover how Chemetrix can elevate your laboratory’s performance.

]]> https://chemetrix.co.za/powerful-partnerships-inspiring-insights/feed/ 0 AFSW2022: The Food Safety Event of the Year! https://chemetrix.co.za/afsw2022-the-food-safety-event-of-the-year-2/ https://chemetrix.co.za/afsw2022-the-food-safety-event-of-the-year-2/#respond Thu, 18 Aug 2022 19:01:56 +0000 https://chemetrix.co.za/?p=4663 Continue reading "AFSW2022: The Food Safety Event of the Year!"]]> Africa Food Safety Workshop 2022 is happening at Emperors Palace in Johannesburg from 27 June until 1 July. The event brings together food safety stakeholders from across the continent to enhance networking and to facilitate the sharing of analytical knowledge and transfer of developed technologies that pertain to all areas of food safety control systems. 

 

The programme

The AFSW2022 programme boasts diverse discussion panels along with influential scientists and key role players delivering presentations that aim to boost systems and enhance trade across the sector. Q&A sessions give attendees the opportunity to engage with the speakers and get additional insights on the exciting range of topics. The event will also have a dedicated exhibition for various Food Safety instrumentation and equipment suppliers, like Chemetrix. 

You can view the full programme here

 

Why should you attend?

With the globalisation of the food chain, there is a greater demand for consistent quality, uncompromising safety and a commitment to protecting consumers while also adopting sustainable solutions. Chemetrix leads the industry with products and services to help you deliver what your customers demand. Our instruments, systems, and supplies are used throughout the food production chain, including incoming inspection, new product development, quality control and assurance, and packaging. We are excited to showcase our products at AFSW2022 along with our expertise, and we look forward to networking with the industry professionals in attendance. 

This event is important to members of various communities connected to food safety, including: 

  • Food testing laboratories (analysts and managers)
  • Quality infrastructure institutions including Accreditation, Standards, Scientific and Legal Metrology institutes
  • Regulators
  • Research/academia
  • Industry (food manufacturers/distributors)
  • International and non-governmental organizations
  • Professional associations

 

The event will be led by Keynote Speaker Dr. Godfrey Bahiigwa (African Union Commission – Director of the Department of Agriculture, Rural Development, Blue Economy and Sustainable Environment (DARBE), AU, Ethiopia).

 

For more event information, visit afsw2022.co.za to purchase your tickets and you can see the discussion and speaker programme here. 

 

]]> https://chemetrix.co.za/afsw2022-the-food-safety-event-of-the-year-2/feed/ 0