ir spectroscopy machine

How to Evaluate an IR Spectroscopy Machine for Your Process

admin 26 min read
How to Evaluate an IR Spectroscopy Machine for Your Process

Screen the measurement architecture before comparing IR spectroscopy machines

Use a measurement-architecture screen before comparing IR spectroscopy machines. The practical questions at this stage are how the sample will reach the optics, which installation context the arrangement must fit, and which instrument trade-offs need confirmation. Organize the comparison around four checks.

  • Sample state and optical access: Record whether the material is gaseous, liquid, or solid, then select a sample holder or interface for that state. An insertion probe measures directly in the process, whereas a flow cell uses an external sample loop. Flow cells are described as easier to condition and suited to conditioned, hazardous, or difficult samples.
  • Installation context: Separate bench, pilot, side-stream, and pressurized use cases. We design the OrionIR® Lab version for recirculation loops off a reactor, side streams on pilot rigs, and bench setups. A high-pressure infrared system can use transmission or attenuated-total-reflection optics in a high-pressure cell to monitor bulk-fluid composition and solid-surface phenomena. The OrionIR® High-Pressure Version uses the same measurement principle as the Lab version, allowing chemometric models built during evaluation to transfer to the pressurized installation.
  • Signal path and instrument type: An IR spectrometer measures changes in infrared radiation after it interacts with a sample. The system generally includes an infrared light source, sample holder, detector, and data-acquisition and analysis software. A trade-off to assess is that dispersive instruments are typically less expensive and have better resolution, while Fourier-transform instruments are more sensitive and faster.
  • Information for supplier review: Prepare the stream and chemistry, expected concentration or composition range, physical state, operating conditions at the interface, measurement location, available samples, reference method, and intended monitoring or model objective. We ask customers to describe the stream and chemistry so we can determine OrionIR® compatibility and configuration.

The output of this screen is a focused comparison brief: a proposed interface, an installation context, and specific questions about calibration material, reference measurements, and application-specific testing. It is not a universal suitability conclusion. Where fit remains uncertain, we offer to test the OrionIR® on customer process samples when it makes sense, giving the next technical discussion a defined configuration and sample context.

Understand what an IR spectroscopy machine measures

An IR spectroscopy machine measures changes in infrared radiation after the radiation interacts with a sample. The resulting spectrum records how the sample affects infrared energy across different wavenumbers or frequencies. The spectrum therefore carries chemical information, but the information still depends on how the sample reaches the optical path and how the data are interpreted.

An IR spectrometer system generally contains four connected parts:

  • An infrared light source emits the radiation that interacts with the sample.
  • A sample holder or interface presents a gas, liquid, or solid to the optical path.
  • A detector captures the radiation after interaction with the sample.
  • Data-acquisition and analysis software processes the detector signal and presents spectral information or a calculated result.

Dispersive instruments separate infrared radiation with a prism or grating and measure the resulting components. Fourier-transform instruments use an interferometer and mathematical transformation to generate a spectrum. That distinction describes how the spectrum is produced; it does not, by itself, decide whether a particular instrument fits your process. Fit still depends on the chemical question, sample presentation, required measurement location, and interpretation method.

The sample interface deserves attention at the same time as the optical principle. A liquid in a recirculation loop, a solid powder in a laboratory holder, and a pressurized process stream impose different physical requirements. If your evaluation defines only the desired analyte and leaves the physical path unspecified, it has not yet defined the measurement. Describe the sample state, the available access point, and the conditions at that point before comparing instruments.

An infographic showing the approximate mid-infrared wavelength range used for molecular vibrations, from 5 microns to 25 microns.
Mid-infrared wavelengths are associated with molecular vibrations.

Separate spectral information from a reported chemical value

A reported chemical value is an interpreted result, not the spectrum itself. Chemometric calibration establishes a mathematical relationship between spectral data and chemical values, so calibration performance is a major determinant of analytical performance. The model uses spectral patterns and reference results to estimate the quantity or property of interest. That relationship must be established for the chemistry and process domain in which you intend to use it.

The distinction matters in practice. A spectrum can change because composition changes, but it can also reflect changes in water content, temperature, physical state, flow, bubbles, solids, or other conditions represented in the process. A model must distinguish the target variation from the variation that affects the measurement without representing the target result. This is why a spectrum-per-second capability does not automatically mean that a concentration result is available at the same rate. Acquisition frequency and analytical interpretation are connected, but they are not interchangeable.

The reference method sets another boundary. We state that predicted values cannot be more accurate than the reference method used for training and that the reference method’s reproducibility determines the model’s accuracy floor. Your evaluation should therefore record the reference method, sample handling, result definition, reproducibility information, and acceptance criteria before model performance is interpreted. If the reference result varies because of its own procedure, the model inherits that limitation.

Use three separate questions when assessing a proposed result:

  1. Does the instrument collect a spectrum that responds to the chemical and physical variation in the sample?
  2. Does the calibration relate that spectral information to the requested chemical value across the intended process domain?
  3. Does the resulting prediction support the decision for which it will be used?

A positive answer to the first question does not settle the second or third. Keep those questions separate in the evaluation brief and in the evidence you request.

Define the process and chemistry the measurement must represent

For fermentation and continuous bioprocesses, an IR spectroscopy machine should be evaluated as part of a measurement path rather than as an isolated instrument. The relevant questions are what stream is observed, when the result becomes available, how the spectrum is related to substrate or product values, and what process decision the result is intended to inform. A suitable configuration still depends on the measurement principle, sample interface, calibration model, reference method, and installation context.

Start with the decision in the run. State whether the application concerns a batch or continuous process, which substrate, product, composition, or process transition matters, and when the result must be available. A batch application may focus on a defined stage or recurring decision. A continuous application may require the measurement location and sample-transport time to be considered alongside the changing stream. The intended use may be process understanding, monitoring, model development, independent validation, or a release-related feasibility question. These purposes require different samples and different ways of reviewing the result.

A useful fermentation measurement plan brings together:

  • The stream, reaction or process stage, and batch or continuous setting.
  • The substrate, product, and other components whose values may vary together.
  • The concentration or composition range expected during the relevant part of the run.
  • The quality question and the process decision that follows the result.
  • The reference method, its result definition, sample handling, reproducibility information, and acceptance criteria.
  • The available samples, batches, feedstocks, and operating states.
  • The proposed measurement location and the route by which material reaches the optical interface.
  • The model objective and the functions responsible for reviewing the result.

This information keeps a fermentation question distinct from a general request for an instrument. For example, a request to follow a substrate during a changing run is not the same as a request to estimate a product value for a later decision. The target value, timing, sample material, reference result, and review process should describe the same application.

An IR spectrometer measures changes in infrared radiation after the radiation interacts with a sample. Dispersive instruments use a prism or grating, while Fourier-transform instruments use an interferometer and mathematical transformation to generate a spectrum. A system generally includes an infrared light source, a sample holder, a detector, and data-acquisition and analysis software. The sample holder or interface must suit the physical state of the material, including whether the sample is gaseous, liquid, or solid.

The instrument label does not by itself determine the fit for a fermentation measurement. A reported comparison describes a trade-off between dispersive and Fourier-transform instruments: dispersive instruments are typically less expensive and have better resolution, while Fourier-transform instruments are more sensitive and faster. The relevant choice depends on the chemical result, sample arrangement, update frequency, and process setting being evaluated.

The spectrum also does not automatically become a substrate or product concentration. Chemometric calibration establishes a mathematical relationship between spectral data and chemical values, so calibration performance is a major determinant of analytical performance. The model therefore needs to be considered together with the material used for calibration and the reference method used to define the chemical result.

For fermentation work, the chemical range matters because the relevant value may change during a run. The calibration set must span the range of expected concentrations, including the extremes. The process material used for calibration should correspond to the conditions the application is intended to represent. Conditions not represented in that material may require additional samples or testing before the model objective can be assessed.

The reference method sets a practical boundary for interpreting the model. Predicted values cannot be more accurate than the reference method used for training, and the reference method’s reproducibility determines the model’s accuracy floor. Before comparing model output with a process target, specify what the reference result means, how the sample reaches that method, how reproducibility is assessed, and which acceptance criteria apply. A numerical output without that context does not answer whether the process question has been addressed.

The route from the fermentation stream to the optics is another central decision. An insertion probe measures directly in the process, whereas a flow cell uses an external sample loop. The selected comparison describes flow cells as easier to condition and suited to conditioned, hazardous, or difficult samples. A direct probe and an external loop therefore represent different arrangements for access, conditioning, maintenance, and sample transport.

For a continuous run, the sample loop can also affect when a measured result represents the process. The comparison lists flow-cell measurement lag as dependent on the sample loop. Include that transport time when deciding whether a result can inform a process transition or control discussion. If the sample is taken from a side stream, describe where it leaves the process, how it reaches the interface, and whether the stream changes during that interval. The measurement location and the intended decision point should not be treated as interchangeable.

The physical state and handling requirements should be stated before selecting the interface. Describe whether the material is a liquid stream, a gas, or a solid-containing sample, and identify the arrangement proposed for presenting it to the optics. Where conditioning is needed, explain which part of the sample path provides it and which process conditions must remain represented in the material used to develop the model. The aim is to connect the sample that reaches the instrument with the stream for which the result will be used.

Configuration questions then become specific to the fermentation installation. Paeonia Innovations designs the OrionIR® Lab version for recirculation loops off a reactor, side streams on pilot rigs, and bench setups. These uses provide a way to frame an evaluation around the available sample path rather than assuming that a laboratory arrangement and a process installation are identical.

A high-pressure application requires its own interface discussion. A high-pressure infrared system can use transmission or attenuated-total-reflection optics in a high-pressure cell to monitor bulk-fluid composition and solid-surface phenomena. Paeonia Innovations’ OrionIR® High-Pressure Version uses the same measurement principle as the Lab version, allowing chemometric models built during evaluation to transfer to the pressurized installation. That stated relationship can be included in the feasibility discussion, while the pressure boundary, cell arrangement, site requirements, and qualified review remain application-specific questions.

For the supplier discussion, describe the stream and chemistry in terms that distinguish the target from the surrounding material. Include the relevant substrate or product, the expected composition range, the material state, the stage or location of measurement, and the batch or continuous context. Note the samples available from the process and which reference results accompany them. If the application includes changing feedstock or product specifications, include those changes because they may affect the calibration material and later model support.

Paeonia Innovations offers the OrionIR® mid-IR spectrometer, which provides one spectrum every second across 900–3500 cm⁻¹ through aqueous, strongly absorbing, and low-concentration streams. The stated capability is an input to an application discussion, not a conclusion that every fermentation stream will produce a suitable model. Paeonia Innovations asks customers to describe the stream and chemistry so it can determine OrionIR® compatibility and configuration.

The OrionIR® interface characteristics can also be framed as questions about the proposed sample path. Paeonia Innovations’ OrionIR® does not use optical fibres, eliminating the limitations from mid-IR transmission and the risk of fibre breakage. Consider how that characteristic relates to the intended installation, while keeping the sample interface, pressure, temperature, access, cleaning, and process responsibilities explicit.

Paeonia Innovations has field-tested the OrionIR® under industrial conditions in North America, South America, Europe, and Asia, and the OrionIR® held its spectral stability across a repeatability test of more than 100 days. Those company-specific results add context to a supplier discussion, but they do not replace testing with the reader’s process samples or establish suitability for an unrelated stream.

A feasibility study can be organized around four questions. First, does the proposed interface present the relevant stream at the point and time associated with the process decision? Second, does the available material represent the substrate, product, composition range, and process conditions the model is expected to address? Third, does the reference method provide a reproducible result against which the spectral model can be interpreted? Fourth, does the proposed laboratory, side-stream, flow-cell, or high-pressure arrangement match the installation being considered?

These questions can be answered with the information available before a supplier discussion. Describe the process stage and decision point, the chemistry and expected range, the sample state, the proposed measurement location, the transport arrangement, the reference method, the samples available for testing, and the intended model use. Identify unresolved items rather than treating a nominal sample, a different process stage, or a different installation as equivalent.

Paeonia Innovations offers to test the OrionIR® on customer process samples when it makes sense. A sample discussion can therefore focus on the stream, chemistry, reference results, sample availability, and proposed measurement arrangement. The purpose is to determine whether compatibility and configuration questions can be examined for the defined application, not to assume a result before testing.

Model support also needs to follow the process rather than end at the initial instrument discussion. Paeonia Innovations offers ongoing chemometric-model support, including revalidation as feedstock, temperature, or product specifications change. Those changes should be identified when the application is described so that the discussion covers how the model may be reviewed as the process changes.

In regulated pharmaceutical work, the site should establish the applicable validation responsibilities and review process with its qualified technical, quality, and regulatory functions. An ICH reflection paper states that ICH Q2 and ICH Q14 address multivariate calibration models associated with analytical procedures, and that process models can support monitoring and control in conventional and advanced manufacturing. The appropriate use, validation approach, and installation requirements remain matters for the responsible site functions.

The practical outcome of this fermentation-focused path is a defined set of questions for the instrument and the process together. If the stream, timing, interface, calibration material, reference result, or installation arrangement remains unclear, that uncertainty identifies the next technical discussion or sample test. If those items are described coherently, Paeonia Innovations can use the stream and chemistry information to discuss OrionIR® compatibility and configuration for the intended process-monitoring application.

Expose the operating envelope before judging suitability

The operating envelope is the set of conditions under which your measurement must remain interpretable. Treat each condition as a fit question rather than a footnote. Then identify which of those conditions are represented in the calibration material and which require additional samples or testing.

This approach changes the question from “Can the instrument measure this chemical?” to “Can the complete measurement system distinguish the required chemical variation across the conditions that will occur?” A liquid sample may be chemically suitable in a controlled laboratory vessel but behave differently in a moving stream. Pressure and temperature may change the process state or the sampling arrangement. These possibilities should become explicit test conditions rather than assumptions.

Create an operating-envelope table for the evaluation:

Condition Expected range or states Represented in calibration material? Additional sample or test needed?
Temperature Process range and transitions Record the covered states Identify gaps
Pressure Normal and exceptional operating states Record the covered states Identify gaps
Composition and concentration Components, ranges, and extremes Record the covered chemistry Identify gaps
Water, solids, and bubbles Expected physical variation Record the covered material Identify gaps
Flow and sample transport Flow states and residence effects Record the covered arrangement Identify gaps
Feedstock and product changes Planned and foreseeable changes Record the covered materials Identify gaps
Instrument configuration Proposed interface and changes Record the evaluated setup Identify gaps

A gap does not establish failure, but it does establish work. Assess which process conditions are represented in the calibration material and obtain additional samples or testing for conditions that are not represented. Suitability should remain conditional until the relevant operating envelope has been examined.

Use the table as a coverage record rather than as a list of specifications. For each row, link the expected state to the material or samples representing it, the reference result available for comparison, and the action required if that state is absent. This keeps the operating envelope connected to the model and the decision instead of treating calibration as a separate activity. If the evaluation includes several stages or operating modes, distinguish them in the record so that the sample set and intended decision remain clear.

A nominal condition should not stand in for an unrepresented transition or composition range. Identify the samples or testing needed before that state is included in interpretation, and record which conditions remain outside the current evaluation. The same record can then support a focused discussion of the proposed measurement arrangement, the calibration material, and the evidence needed for the intended process question. It also provides a basis for reviewing changes to the process or configuration without silently extending the original conclusion.

Choose how the sample reaches the optics

The sample interface determines how material reaches the optics, so interface selection is part of the analytical method. Select the sample holder or interface for the sample’s physical state, including whether the sample is gaseous, liquid, or solid. Then map the measurement location, transport path, pressure and temperature conditions, conditioning needs, access for maintenance, and the effect of the interface on timing.

The selected comparison describes flow cells as easier to condition and suited to conditioned, hazardous, or difficult samples. A direct probe may reduce transport between the process and the optical measurement, while a flow-cell arrangement may give you more control over conditioning and access. Neither description decides the fit for your application; the process and site conditions do.

Use a direct-interface assessment when these questions matter:

  • Can the probe reach a representative location without creating an unacceptable process intrusion?
  • Will the local temperature, pressure, solids, bubbles, and flow conditions be within the proposed arrangement?
  • Can the interface be accessed for cleaning, inspection, and maintenance?
  • Does the direct location represent the material involved in the decision?

Use a flow-cell assessment when these questions matter:

  • What sample loop, bypass, or side-stream arrangement transports material to the cell?
  • What transport time or lag separates the process event from the measurement?
  • What conditioning is needed before the sample enters the optical path?
  • How will pressure, temperature, cleaning, hazardous material, and access be managed at the cell?

A flow cell is not simply a remote version of a probe, and a probe is not automatically the simpler option. Document the complete material path and include it in calibration and testing. If the tested interface differs from the proposed installation, treat transfer between them as an open evaluation question.

Match the configuration to the installation context

The installation context determines which configuration questions must be answered. Separate a bench or laboratory evaluation, a recirculation loop or pilot side stream, and a pressurized process installation instead of treating them as interchangeable settings. Compare the sample path, pressure boundary, optical arrangement, access, cleaning, data connection, and operating responsibilities for each proposed context.

A laboratory or pilot evaluation can use a controlled sample arrangement while you establish the chemical relationship and assess process samples. A controlled sample arrangement provides an evaluation path when process material can be made available outside the primary vessel or line. It does not establish suitability for every stream or installation.

A high-pressure application adds a pressure-rated measurement context. A high-pressure infrared system can use transmission or attenuated-total-reflection optics in a high-pressure cell to monitor bulk-fluid composition and solid-surface phenomena. Your assessment should therefore identify the pressure conditions, cell arrangement, optical mode, material path, and site controls required for the intended measurement. The existence of a high-pressure optical arrangement does not remove the need to test the actual chemistry and operating envelope.

For the OrionIR® High-Pressure Version, we state that the measurement principle is unchanged from the Lab version, allowing chemometric models built during evaluation to transfer to the pressurized installation. Treat that as a configuration-specific supplier input to examine during the evaluation, not as a universal transfer result. Model transfer still depends on whether the evaluation samples, interface, pressure conditions, and process domain represent the final application.

Record installation questions that specifications alone may not answer:

  • Where will the instrument and sample interface be located?
  • What pressure, temperature, cleaning, and access conditions apply?
  • Which power, data, control-system, and integration requirements must the site meet?
  • Which installation, validation, safety, and operations functions own each decision?
  • Which parts of the proposed configuration will be reproduced during testing?

Use the configuration that matches the physical and decision context you intend to operate, and keep any untested transfer as a defined boundary.

Treat calibration material as part of the instrument decision

Calibration material is part of the instrument decision because the model can only learn the chemical and process relationships represented by its samples. Include the expected concentration range, including the extremes, rather than building a set around only typical material.

Our calibration set must span the range of expected concentrations, including the extremes. The same principle applies to relevant process variation: if temperature, water content, feedstock, product specification, solids, bubbles, or flow conditions can change the measured relationship, the evaluation should determine whether those states are represented. A collection of convenient laboratory samples may be useful for initial work, but it may not represent the operating domain that the model must interpret.

Plan the calibration set around variation that can affect the intended result:

  • Include samples across the expected low-to-high concentration or composition range.
  • Include material from relevant process stages, batches, feedstocks, or continuous operating states.
  • Record temperature, pressure, flow, water content, solids, bubbles, and other conditions associated with each sample.
  • Keep sample handling and reference-method procedures consistent with the intended comparison.
  • Mark conditions that are absent, rare, or outside the planned evaluation range.
  • Define what additional sampling or testing is required before a missing condition can be included in the model domain.

Calibration is not only a software activity. The sample set, reference method, interface, and process conditions all contribute to the relationship the model uses. If you change the feedstock, temperature, product specification, sample path, or instrument configuration, assess whether the existing calibration still represents the measurement. A supplier may support further testing or model work, but the particular change and its effect must be defined for your application.

The practical decision is whether the calibration material supports the process question you need answered. If it does not, expand the material or narrow the intended claim until the missing conditions have been addressed.

Set the reference-method boundary

The reference method defines what a model prediction can reasonably mean. Before evaluating an IR system, document the process setting, chemistry, quality question, reference method, sample set, measurement conditions, data and model objective, and governance responsibilities. Include how the reference result is produced, how samples are handled, how reproducibility is assessed, and which acceptance criteria apply.

This boundary has a direct consequence: an instrument evaluation should not compare a predicted value with an undefined or unstable reference value and then treat the difference as an instrument characteristic. The comparison method must be described before its results are used to judge a model.

Separate the following elements in your evaluation record:

  • The chemical or process quantity that the reference method measures.
  • The sample identity, collection point, handling, storage, and preparation.
  • The procedure used to produce the reference result.
  • The reproducibility or expected variation of that procedure.
  • The range, acceptance criteria, and decision rule applied to the result.
  • The relationship between the reference measurement and the IR spectrum collected from the same or comparable material.

This separation also helps you identify a mismatch between the intended decision and the available reference. A model for monitoring a changing process may require samples that cover process transitions, while a laboratory method comparison may use a narrower and more controlled set. A release-related question may require a defined governance path and independent validation beyond an exploratory model.

Do not describe a model as accurate in isolation. State what it predicts, against which reference, across which range and conditions, and for which decision. If the reference method does not support the intended interpretation, improve the comparison basis or narrow the proposed use before drawing a conclusion about the instrument.

Validate with independent process evidence

Independent process evidence tests whether the calibrated relationship remains useful beyond the samples that created it. Independence is part of the evidence; a second measurement of the same calibration population does not answer the same question.

For pharmaceutical NIR method validation, the selected review states that validation samples should come from completely different batches and use at least one different API lot than the calibration sample. That statement belongs to the pharmaceutical regulatory context and should be applied within the method and governance requirements for the particular application. It should not be generalized to every industry or every IR use.

Build the validation plan around the intended process domain:

  • Reserve samples that were not used to construct the spectral library or calibration model.
  • Cover the calibration range, including relevant low and high conditions.
  • Include process variability that you expect during commercial or routine operation where possible.
  • For pharmaceutical applications, document batch and API-lot independence where the applicable method-validation approach requires it.
  • Compare predicted and reference results using the acceptance criteria defined for the decision.
  • Record conditions under which the model produces an out-of-range, low-quality, or otherwise non-reportable result.

Validation should also examine the boundary of the method. Assess whether samples outside the method’s scope are rejected when spectral quality thresholds are not met or predictions fall outside the model range. This question matters because a result that looks numerical can still be outside the conditions for which the model was developed.

An independent validation set does not prove universal suitability. It provides evidence for the tested chemistry, sample interface, process conditions, reference method, and decision. State those limits beside the conclusion so that the evidence is not carried into a different process, configuration, or jurisdiction without a separate assessment.

Connect the model to the manufacturing decision

State whether your objective is process understanding, monitoring, control, model development, independent validation, or a release-related feasibility question. Then identify who reviews the result, what happens when it passes or fails, and which evidence is required before the result can influence the process.

The selected ICH reflection paper states that ICH Q2 and ICH Q14 address multivariate calibration models associated with analytical procedures, and that process models can support monitoring and control in conventional and advanced manufacturing. The same paper describes applications across conventional and advanced approaches, including continuous manufacturing. This provides a regulatory and manufacturing context for considering models, but it does not decide the requirements for your site, product, jurisdiction, or intended use.

Connect the model to operations through a written decision map:

  1. Define the result and the process question.
  2. Identify the reference method and the model domain.
  3. Specify the conditions under which the result is reportable.
  4. Define the response to an in-range result, an out-of-range prediction, poor spectral quality, or missing data.
  5. Assign review and approval responsibilities across process development, quality, operations, engineering, and regulatory functions as applicable.
  6. Record how changes to feedstock, temperature, product specification, sample interface, or model will be assessed.

This map keeps a monitoring result from being treated as an automatic control action or release decision. It also makes the evaluation more concrete: you can ask whether the proposed data rate, interface, model, and validation evidence support the decision you have identified.

Model governance continues after installation. We offer ongoing chemometric-model support, including revalidation as feedstock, temperature, or product specifications change. The need for that support should be tied to the changes and decision rules in your application, rather than described as a general assurance.

Use OrionIR® capabilities as supplier-fit questions

OrionIR® characteristics should enter your evaluation as supplier-fit questions tied to your stream and decision. We offer the OrionIR® mid-IR spectrometer, which provides one spectrum every second across 900–3500 cm⁻¹ through aqueous, strongly absorbing, and low-concentration streams. Those are the documented characteristics to examine against your chemistry, concentration range, sample interface, and required result. They are not a universal conclusion that every stream or application will produce a usable model.

Ask how the stated spectral range and acquisition rate relate to your application:

  • Which chemical features in your stream fall within the proposed spectral range?
  • What concentration and composition conditions will be represented in the evaluation?
  • Does one spectrum every second provide the information needed for your process decision, after sampling, transport, preprocessing, and model interpretation are considered?
  • How will water, strong absorption, low concentration, solids, bubbles, or changing flow affect the tested measurement?
  • Which interface and configuration will reproduce the intended installation?

We use the stream and chemistry you provide to determine OrionIR® compatibility and configuration. Use that request to provide the process brief rather than treating compatibility as a label based only on the analyte name. We also state that the OrionIR® does not use optical fibres, eliminating the limitations from mid-IR transmission and the risk of fibre breakage. Keep that statement limited to the documented instrument construction; it does not replace an assessment of the sample path, optical access, process conditions, or installation requirements.

We state that the OrionIR® held its spectral stability across a repeatability test of more than 100 days. Treat this as a reported repeatability-test result for the instrument and ask how stability relates to your proposed measurement, calibration, maintenance, and revalidation plan. A stability result does not establish the accuracy of a chemical prediction or the suitability of an untested process.

The supplier-fit question is therefore specific: can the documented OrionIR® configuration, tested with your material and reference method, support the result and decision defined in your brief?

A branded infographic showing the OrionIR® documented spectral range of 900–3500 cm⁻¹ and a reported spectral stability repeatability test of more than 100 days.
These supplier-stated characteristics still require testing with the intended material and reference method.

Prepare a focused feasibility discussion

A focused feasibility discussion starts with a complete measurement brief. Give the discussion enough process detail to examine the instrument, sample interface, calibration material, reference method, and intended decision together.

Prepare these items before contacting the team:

  • Process identity: Name the reaction, process stream, stage, batch or continuous setting, measurement point, and intended decision point.
  • Chemistry: List relevant components, expected concentration or composition ranges, water content, solids, bubbles, and material characteristics that may affect the measurement.
  • Operating conditions: Record temperature, pressure, flow, cleaning conditions, sample transport, access, and any classified-area or site requirements that apply.
  • Measurement arrangement: Describe a proposed probe, flow cell, bypass, side stream, recirculation loop, or laboratory setup, including transport time and conditioning.
  • Reference method: Identify the comparison method, result definition, sample handling, reproducibility information, and acceptance criteria.
  • Samples: State which process samples, batches, feedstocks, operating states, and concentration extremes are available for testing.
  • Model objective: Specify whether the work concerns process understanding, monitoring, model development, independent validation, control, or a release-related question.
  • Governance: Identify the functions that will review the result and describe the action associated with an in-range, out-of-range, failed, or unavailable result.
  • Change conditions: List planned changes to feedstock, temperature, product specification, process scale, interface, or instrument configuration.

We offer to test the OrionIR® on customer process samples when it makes sense. Sample testing should answer a defined feasibility question, such as whether the proposed interface collects usable spectra across the intended process domain or whether the reference method and sample set support a model for the stated result. It should not be treated as a demonstration of universal suitability.

Send a concise brief with the available sample and process information, then identify the unresolved technical questions. That gives the supplier discussion a clear starting point and leaves the conclusion tied to the application you actually intend to measure.

State what the evaluation cannot establish on its own

An instrument evaluation cannot establish universal suitability from a single spectrum, sample, or configuration. A positive result applies to the chemistry, process conditions, interface, calibration material, reference method, and decision that were actually examined. Do not treat an exploratory measurement as proof that the instrument will replace an existing analytical method, support a release decision, or perform the same way after a material process or configuration change.

Keep the following boundaries visible in the evaluation record:

  • Conditions absent from the calibration material require additional samples or testing before they are included in the model domain.
  • Validation samples should be entirely independent of the samples used to build the spectral library and should cover the calibration range and, where possible, process variability.
  • Samples outside the method’s scope should be assessed for rejection when spectral quality thresholds are not met or predictions fall outside the model range.
  • A model result remains bounded by the reference method, its reproducibility, the sample handling, and the tested range.
  • Transfer from a laboratory, pilot, side-stream, or high-pressure arrangement requires attention to the differences between the evaluated and intended configurations.

This boundary is especially relevant when the process changes. Feedstock, temperature, product specifications, pressure, water content, solids, flow, sample conditioning, and instrument configuration can alter the conditions represented by the model. A change may require additional samples, testing, model work, or revalidation rather than a simple assumption that the previous conclusion still applies.

Use the evaluation to reach a stated, conditional conclusion: identify what was tested, what result it supported, what conditions were covered, and what work remains before the result can inform the intended process decision. The conclusion is useful when its scope is clear.

Take the next technical step toward application fit

Your next technical step is to turn the process question into a measurement brief covering the stream, chemistry, concentration range, operating conditions, sample interface, reference method, available samples, model objective, and decision governance. The stream and chemistry details in that brief give us the inputs to determine OrionIR® compatibility and configuration. We offer to test the OrionIR® on customer process samples when it makes sense, and we offer ongoing chemometric-model support, including revalidation as feedstock, temperature, or product specifications change.

A qualified technical, quality, safety, or regulatory professional should determine individual suitability and the requirements for your site. Send the completed process brief when you are ready to discuss the application and its evidence needs.



Next step

Have a question for Paeonia Innovations?

Get in touch