mid infrared spectrometer
How to Evaluate a Mid-Infrared Spectrometer for Real-Time Bioprocess Monitoring
Is a mid-infrared spectrometer a fit for your process-monitoring question?
A mid-infrared spectrometer can be a plausible fit for real-time monitoring of multiple analytes in complex fermentation fluids, but the fit depends on the measurement question, the sample path, and the evidence available for your process. Mid-infrared spectroscopy has been evaluated for real-time measurements made either in situ or ex situ, while deployment barriers and application limitations remain. Treat the instrument as part of a measurement system rather than as an isolated specification.
Start by naming the result you need. Your target might be a substrate concentration, a product concentration, or another defined process measurement. Then state how that result will be used: for observation, an operating decision, process optimization, or a control action. A supplier can assess a proposed arrangement more meaningfully when the requested result and its intended use are explicit.
Your initial fit question should cover five areas:
- Measurement: Which substrate, product, or other process variable must be estimated, and over what expected concentration range?
- Material: What is the stream composition, and does it contain solids, multiple phases, or constituents that may produce overlapping spectral features?
- Conditions: What temperature, pressure, and sampling location will apply during the measurement?
- Reference: Which laboratory or process method will provide the values used to establish and assess the model?
- Deployment: Will the material reach a bench setup, a pilot side stream, a recirculation loop, an extractive system, or a pressurized installation?
The practical decision is therefore not simply whether mid-infrared spectroscopy can produce a spectrum. The decision is whether your stream, measurement conditions, reference data, sampling arrangement, calibration coverage, and validation plan can support an interpretable process result. Build the supplier-test brief around those questions before selecting a permanent configuration.
A concentration estimate requires more than a continuous spectral stream. It requires a reference basis, a suitable multivariate or chemometric model, and evidence that the model addresses the intended process conditions. PLS models are empirical, so changes in process conditions, raw materials, analyte relationships, or instrumentation can reduce performance when those changes are not represented in calibration.
That boundary gives you a useful first decision rule: request a technical assessment when you can describe the target result and stream, but do not treat an initial spectrum or a successful laboratory correlation as sufficient evidence for every operating condition. The rest of this guide turns that rule into a structured evaluation.
What a mid-infrared spectrometer measures
A mid-infrared instrument exposes a sample to infrared radiation and analyzes the fraction absorbed at specific frequencies. An FTIR instrument records the spectrum through an interferometer-based approach rather than treating each frequency as an entirely separate measurement. The resulting spectrum is a pattern of absorbance across a range of infrared frequencies.
That pattern carries chemical information because different constituents interact with infrared radiation at different frequencies. The instrument records the response; it does not automatically label each feature as a substrate or product concentration. Interpretation depends on how the spectrum relates to reference measurements from the process or an appropriate laboratory method.
For a fermentation or continuous bioprocess, this distinction matters. A spectrum can change as the broth changes, but the direction and size of that change do not by themselves establish the concentration of a named analyte. You need a defined measurement target and a model built for the material and conditions in which you intend to use it.
The Beer–Lambert law relates absorbance to the concentration of the absorbing species and the optical path length, providing a basis for quantitative spectroscopic measurement. In a simple system, that relationship can help connect an optical response with concentration. Complex fermentation streams require more care because several constituents may absorb in related regions, the matrix may be inhomogeneous, and optical properties may vary.
The ATR-FTIR evidence shows why a direct calculation is not always sufficient for a complex matrix: inhomogeneity, variable optical properties, and heavily overlapping absorption bands can make direct quantitative analysis uncertain. Beer–Lambert calculations alone may not resolve overlapping peaks. For your evaluation, ask how the proposed measurement arrangement and model will address the actual stream rather than assuming that a visible spectral change is an independently validated concentration result.
Overview
A mid-infrared spectrometer can contribute to real-time bioprocess monitoring when three parts of the measurement chain align: the process question is defined, the sample reaches the measurement in a controlled way, and the spectral data are connected to suitable reference measurements through a model. Mid-infrared spectroscopy can support real-time, in situ or ex situ monitoring of multiple analytes in complex fermentation fluids, but deployment limitations remain.
Use the following three-part screen before comparing instrument specifications:
- Defined result: You can name the analyte or process measurement, explain its intended use, and state the concentration range or operating range that matters.
- Represented stream: You can describe the chemistry, temperature, pressure, solids, sampling arrangement, and other conditions that the proposed system must accommodate.
- Supported interpretation: You can identify the reference method, calibration samples, validation approach, residual monitoring, documentation, and operating checks that will support the displayed result.
If one part is unclear, the next step is to resolve that uncertainty rather than to assume that a different instrument label will solve it. A supplier-test brief helps separate questions about optical measurement from questions about sampling, model construction, and deployment.
A continuous spectral stream can support real-time monitoring and process optimization, but a concentration estimate requires a reference basis, a suitable multivariate or chemometric model, and evidence covering the intended process conditions. PLS models are empirical and can lose performance when unforeseen changes, analyte correlations, or process, raw-material, and instrument variation fall outside the calibration conditions.
This means that a mid-infrared spectrometer may be useful for more than one stage of development, but the evidence required at each stage is not interchangeable. A bench evaluation can answer questions about sample interaction and model feasibility. A pilot or pressurized assessment must also address the installation context and relevant operating variation. A process result should not be treated as established merely because the instrument produced repeatable spectra in a different arrangement.
From spectral data to concentration estimates
A concentration estimate can be produced by linking spectra with reference measurements through a model that represents the intended application. The measurement chain can be described in four stages:
- Acquire: The instrument records the infrared response from the sample and produces a spectrum.
- Pair: The spectrum is associated with a reference value obtained through the selected reference method.
- Model: A multivariate or chemometric model learns the relationship between the spectral data and the reference result across representative samples.
- Assess: The model is tested under conditions that show whether its outputs remain interpretable for the intended process.
The reference basis is not a formality. It defines what the displayed value means and gives the model a target against which it can be built and assessed. The model also needs coverage of the material and conditions that matter for use. If a process changes in ways that were not represented during calibration, the relationship learned by the model may no longer describe the new samples.
The concentration output depends on the quality and relevance of the reference data, the selected modelling approach, and the application conditions represented during calibration and validation.
PLS calibration models are empirical and should be established under application-relevant conditions before use. Unforeseen changes can reduce performance, as can correlations between the analyte of interest and other constituents when those relationships are not adequately represented. Process variation, raw-material variation, and instrument variation can create the same problem.
For your process, ask which variation the calibration set represents and which variation remains outside it. A model that distinguishes the target analyte only because the analyte changes together with another constituent may not behave as expected when that relationship changes. A model that works on one instrument or in one arrangement may also require evidence before you rely on it in another arrangement.
The useful output of this stage is not a single fit statistic in isolation. It is a documented explanation of the reference method, sample population, operating conditions, model scope, validation samples, and boundaries on interpretation.
The reference method sets a practical boundary on the model output. Our predicted values cannot be more accurate than the reference method used for training; the model’s accuracy floor is determined by that method’s reproducibility. That statement applies directly to how you should interpret a proposed calibration plan.
Ask how reference samples will be collected, how the reference result will be established, and how reproducibility will be characterized. If the reference measurement is inconsistent, the spectral model cannot remove that uncertainty simply by processing more spectra. The model can represent the relationship available in the training data, but it cannot create a more reproducible reference basis than the method provides.

Key Highlights
A sound evaluation connects the process question to the stream, the sample path, and the model evidence. Keep these practical points in view:
- Define the target substrate, product, or other process measurement before comparing instruments.
- Describe stream composition, concentration range, temperature, pressure, solids, sampling arrangement, and intended accuracy requirement.
- Identify the reference method and obtain calibration samples that represent the expected range, including its extremes.
- Treat validation, residual monitoring, documentation, and operating checks as part of the measurement plan.
- Separate a spectrum from the concentration estimate created by a model.
- Treat a laboratory result as evidence for the conditions tested, not as automatic evidence for every pilot, pressurized, or plant condition.
- Ask whether the proposed configuration and model address the specific stream rather than relying on a general application label.
Configuration and interpretation depend on the material and operating context.
The immediate action is to turn your process information into a short supplier-test brief. Include the result you need, the samples and reference data you can provide, the expected operating range, and the installation context. Use the brief to expose unanswered technical questions before you select a permanent instrument arrangement.
Build the supplier-test brief around the stream
A supplier-test brief should describe your stream and measurement question before it describes a preferred instrument. The purpose is to let a technical team map the chemistry, sampling conditions, calibration evidence, and installation needs to a proposed evaluation. Comparing spectral ranges or instrument names in isolation leaves the most consequential fit questions unanswered.
Write the brief in the following order:
- Process objective: State whether you need a substrate concentration, product concentration, or another defined process measurement. Explain what decision the result will inform.
- Stream identity: Describe the main constituents, solvent or aqueous character, phases, solids, and any expected changes during the run.
- Concentration range: Give the expected lower and upper values for each target and identify the extremes that the calibration must represent.
- Operating conditions: Record temperature, pressure, flow or recirculation context, and whether those conditions change during the run.
- Sampling location: Identify the reactor loop, pilot side stream, bench setup, extractive line, or proposed plant location from which the material will reach the measurement.
- Reference basis: Name the reference method, how samples will be collected, and which reference results are available.
- Evidence request: Ask how calibration, validation, residual monitoring, documentation, operating checks, and model maintenance will be handled.
This order keeps the process question in view. It also gives you a written basis for deciding whether a proposed test answers the question you actually need to answer.
The concentration range deserves specific attention because the calibration set must span the expected range, including the extremes. A set built only around convenient or typical samples may not represent the low and high conditions that matter for your process decision. State the expected range before sample selection, then ask how the available samples cover it.
Include unusual but credible conditions in the brief when they are part of the intended use. A change in raw material, process stage, solids content, or operating condition may alter the relationship between the spectrum and the reference result. The supplier discussion should identify whether those conditions belong in the evaluation, in a later validation exercise, or outside the proposed model scope.
For an OrionIR® discussion with us, describe the stream and its chemistry so we can determine compatibility and configuration. Provide the information in a form that distinguishes known conditions from estimates, and identify which samples and reference results can support testing. A clear brief gives the technical discussion a defined starting point without assuming that OrionIR® is suitable for every fermentation or continuous-bioprocess application.
Check matrix complexity and operating conditions
Matrix complexity affects how you should assess a proposed mid-infrared measurement. Complex streams can contain several absorbing constituents, variable optical properties, and nonuniform material. The ATR-FTIR evidence shows that direct quantitative analysis can be uncertain under those conditions, and that Beer–Lambert calculations alone may be insufficient when absorption bands overlap.
Fermentation broth introduces another relevant condition: fermentation broth is mostly water, and water absorbs strongly across large parts of the mid-infrared spectrum. That does not determine the outcome for every process, but it makes the aqueous matrix part of the technical question. Ask how the proposed sampling arrangement, optical path, spectral information, and model development address the actual broth rather than treating water as an incidental background.
Record the following matrix details in your brief:
- Main components and expected by-products or additives.
- Whether the material is homogeneous, particulate, multiphase, viscous, or changing during the run.
- Expected concentration ranges for each target and relevant non-target constituents.
- Temperature and pressure at the proposed measurement point.
- Solids content and any condition that may change the sample presented to the measurement.
- The accuracy requirement and the reference method that will define the comparison.
These details do not establish suitability by themselves. They define the conditions that the supplier test and model evidence must address.
Use matrix information to choose the next evaluation question. If composition or phase behavior is uncertain, prioritize representative sample testing and a discussion of the sampling interface. If overlapping bands or variable optical properties are expected, ask how the model will distinguish the target from related constituents. If pressure or temperature changes during operation, include those conditions in the installation and validation discussion.
Do not convert the presence of a strong spectral response into a concentration claim without the reference basis and model evidence that connect the response to your target. The appropriate boundary is to test the specific stream under conditions relevant to the intended use.

Define how the sample reaches the measurement
The sample interface determines what material the instrument actually measures and under which conditions. One process FTIR configuration continuously extracts a sample from the process stream, directs it to the analyzer, and returns the sample to the process after analysis. That extractive arrangement is one possible system architecture, not a universal requirement for every bioprocess.
Our OrionIR® Lab version is designed for recirculation loops off a reactor, side streams on pilot rigs, and bench setups. Those arrangements give you specific starting points for an evaluation, but they do not remove the need to examine stream chemistry, pressure, temperature, solids, and the route from the process to the measurement.
Describe the sample path in concrete terms:
- Where the sample leaves the process.
- Whether it moves through a recirculation loop, side stream, bench vessel, or extractive line.
- Whether the sample returns to the process after measurement.
- Which pressure and temperature conditions apply at the measurement point.
- How representative the measured material is of the process result you need.
- Which changes in flow, composition, or phase behavior could alter the sample presented to the instrument.
A supplier-test brief that omits the sample path cannot distinguish an optical question from a process-interface question. Define the path before treating a laboratory spectrum as evidence for another installation.
Choose the evaluation arrangement that reflects the decision you need to make. Use a bench or pilot side-stream arrangement when that is the available route for representative material, and document how it relates to the intended process location. Consider an extractive arrangement when the proposed system requires the sample to be transported to the analyzer, and examine the return path as part of the system description.
The direct action is to draw the sample route, list its conditions, and ask the supplier to assess that route. Do not assume that a configuration designed for a recirculation loop or bench setup automatically establishes suitability for a different installation.
Separate laboratory evaluation from pressurized installation
A laboratory result and a pressurized installation answer related but different questions. Our OrionIR® High-Pressure Version uses the same measurement principle as the Lab version, allowing chemometric models built during evaluation to transfer to a pressurized installation. That is a product-specific configuration statement.
The broader model-transfer question remains application-specific. PLS models are empirical, and instrument-related, process, raw-material, or other variation can reduce performance when it is not represented during calibration. The same measurement principle therefore does not, by itself, establish that a model will remain interpretable under every pressure, temperature, sampling, or process condition.
Keep the two evidence questions separate:
- Measurement and configuration: Does the proposed version provide the intended measurement arrangement for the pressure and sample path?
- Model transfer: Does the calibration and validation evidence show that the model remains suitable when the installation context changes?
- Process representation: Do the samples and reference results cover the material and operating variation expected at the pressurized location?
- Operating controls: Are documentation, residual monitoring, and operating checks defined for the intended use?
Ask for evidence against each question rather than treating the product configuration statement as a substitute for application-relevant validation.
If your evaluation uses a laboratory arrangement and the intended installation is pressurized, carry the same target measurement and representative material into the transfer assessment, then test the relevant process and instrument variation. The supported boundary is that our High-Pressure Version uses the same measurement principle as the Lab version; broader reliability still depends on application-relevant calibration and validation.
If your process will remain at laboratory or pilot scale, define that scope clearly and avoid presenting the result as evidence for an untested plant condition. If a plant installation is the intended destination, include its pressure, temperature, sampling location, operating changes, documentation, and checks in the supplier-test brief before relying on a transferred model.
Test the reference method and calibration coverage
The reference method defines the result against which the spectral model is trained and assessed. For that reason, the reference method belongs in the first version of your supplier-test brief, not as an afterthought once spectra have been collected.
Describe how reference samples will be selected, handled, and measured. State which process conditions each sample represents and how the result will be paired with the corresponding spectrum. If the process contains several targets, identify the reference method for each target rather than assuming that one laboratory result describes every model output.
A useful reference-method discussion covers:
- The measurement principle and reported units of the reference method.
- The expected reproducibility of the reference result.
- The sample collection point and timing relative to spectral acquisition.
- The range of reference values available for calibration and validation.
- The process conditions represented by those samples.
- Any conditions under which the reference method may not represent the material measured by the spectrometer.
The aim is not to make the reference method sound more precise than it is. The aim is to establish what the model is being asked to reproduce and where its interpretation must stop.
Calibration coverage should span the expected concentration range, including the extremes. Coverage should also reflect the application-relevant conditions that may change the spectral relationship, including process variation, raw-material variation, and instrument or installation variation where those conditions belong to the intended use.
Separate three questions in your test plan:
- Range coverage: Do the calibration samples include the lower and upper expected concentrations and the conditions between them?
- Condition coverage: Do the samples represent the stream composition, temperature, pressure, solids, and process states in which the output will be used?
- Model assessment: Are validation samples and operating checks available to test performance outside the data used to establish the model?
A calibration set that covers concentration but not the relevant process variation may leave an important source of uncertainty unaddressed. Ask the supplier to identify the scope represented by the samples and to state which conditions require additional evidence.
Our calibration set must span the range of expected concentrations, including the extremes. When you prepare an evaluation request for OrionIR®, list the expected lower and upper values, identify the samples available at those boundaries, and state where reference-method reproducibility may limit interpretation. That information gives the calibration discussion a defined technical basis.
Plan validation and model maintenance
Validation should show whether the model remains interpretable for the intended process, not merely whether it describes the samples used to build it. The evaluation plan should address validation samples, residual monitoring, documentation, and operating checks alongside the target measurement and sampling arrangement.
PLS models are empirical and can degrade when unforeseen changes occur or when analyte correlations, process variation, raw-material variation, or instrument variation are not represented in calibration. Real-time bioprocess monitoring also requires application-specific calibration for spectroscopic techniques such as UV-Vis, NIR-MIR, Raman, and fluorescence. The calibration requirement is therefore part of the application, not a generic property that can be transferred unchanged from one process to another.
Define validation around the conditions that matter to you:
- Use samples that were not used to establish the model when assessing its behavior.
- Include relevant process states, concentration ranges, and material variation.
- Record residuals or other agreed checks that can identify a departure from the model’s represented behavior.
- Document the model scope, reference method, sample conditions, installation arrangement, and operating assumptions.
- Establish which operating checks are required before and during use.
These are evaluation requirements, not promises of a particular accuracy or outcome. Their purpose is to make the boundary of the displayed result visible.
Keep calibration, validation, and operating checks separate when reviewing the evidence. Calibration establishes the relationship between spectra and reference results for the represented samples. Validation tests whether that relationship remains interpretable on samples or conditions not used to establish it. Operating checks monitor whether subsequent measurements remain within the documented scope of the model and installation.
A proposal is easier to assess when it identifies which samples and conditions support each role. Otherwise, a strong relationship within the calibration set can be mistaken for independent evidence that the model will remain interpretable during use. For each target, ask the supplier to state the represented concentration range and process conditions, which conditions were held out for validation, and what observation would prompt new reference sampling or model review. This is not a universal acceptance threshold; it makes the decision boundary explicit and prevents an unexplained residual from being treated automatically as a confirmed concentration change.
If the stream, installation arrangement, or operating range changes after the evaluation, return to the brief and ask whether the original calibration and validation evidence still covers the new use. The answer should be based on the changed conditions and reference evidence, not simply on the fact that the same instrument continues to produce spectra.
Model maintenance begins with a plan for recognizing changed conditions. Ask what happens when a raw material changes, the process moves outside the original concentration range, the sample interface changes, or the instrument is installed in a different context. Ask which new reference measurements would be needed and how residual or operating checks would trigger an assessment.
Do not treat a model as complete because an initial calibration produced a useful relationship. Treat it as bounded by the conditions represented in the calibration and validation evidence. If the intended use changes, the technical discussion should identify whether additional samples, model development, validation, documentation, or operating checks are required before relying on the output.
Distinguish independent evidence from OrionIR® capabilities
Independent evidence supports the general proposition that mid-infrared spectroscopy can contribute to real-time monitoring of multiple analytes in complex fermentation fluids, while also identifying deployment barriers and limitations. A review published on December 2, 2025 reported that vibrational spectroscopy, including MIR and Raman, provides molecular fingerprinting and that Raman can be particularly effective in aqueous systems because of minimal water interference.
Those observations help frame an instrument decision, but they do not select a technique for your process or establish a model for your stream. The same review identified application-specific calibration as a requirement for spectroscopic techniques used in real-time bioprocess monitoring. Use independent evidence to understand the measurement context, then ask for application evidence tied to your material, target, sampling arrangement, and operating conditions.
A near-infrared fermentation-monitoring study reported partial-least-squares models for lactic acid, sophorolipids, and sodium gluconate, with R2 greater than 0.98 for the tested systems. That result belongs to the tested near-infrared systems, analytes, samples, and conditions. It should not be transferred to a mid-infrared model, a different analyte, or your process without application-specific evidence.
The comparison is useful because it shows why technique-level claims and application-level claims must remain separate. A reported result from another spectral region can inform the questions you ask about calibration and validation. It cannot serve as evidence that a proposed mid-infrared configuration will produce the same result for your fermentation or continuous process.
OrionIR® capabilities should be stated separately from independent method evidence. General evidence can establish that mid-infrared spectroscopy is a relevant approach for real-time bioprocess monitoring. Product information can describe a proposed OrionIR® measurement rate, spectral range, or configuration. Neither category removes the need to connect the model to your reference method and intended process conditions.

Map the brief to a potential OrionIR® configuration
Compare the proposed sample path and operating pressure with the available configuration information. The available configuration information does not establish universal suitability for every process.
Map your brief to the configuration questions below:
- Which version corresponds to the proposed sample path and operating pressure?
- Can representative material reach the evaluation arrangement?
- Which temperature, pressure, solids, and composition conditions must be represented during testing?
- Which reference samples and concentration extremes will support the model?
- Which tests will examine model interpretation across the conditions represented in the evaluation?
Bring the completed process description, rather than a product name alone, to the technical discussion.
Use the OrionIR® information as a bounded input to the fit decision. The stated spectral range, spectral rate, Lab arrangement, and High-Pressure Version provide questions for configuration assessment. They do not replace evidence about your target measurement, reference method, calibration coverage, validation, or installation conditions.
Know when to request testing, modelling, or an instrument loan
Request sample testing when the stream chemistry, matrix complexity, sampling interface, or spectral response remains uncertain. Sample testing is especially relevant when you need to determine whether representative material can reach the proposed arrangement or whether the target measurement can be connected to reference results under the intended conditions.
Prepare the samples and information needed for a meaningful assessment:
- Identify the target analytes or process measurement.
- Describe the stream and its chemistry, including expected concentration range and extremes.
- State temperature, pressure, solids, phases, and the proposed sample path.
- Provide the reference method and any paired reference results available.
- Explain which process states the samples represent.
End this branch with a defined request: ask for an assessment of the supplied samples against the target measurement and proposed configuration. Do not treat a test on an unrepresentative sample as evidence for the full process.
Request chemometric model development when you have a defined target but need to establish the relationship between spectra and reference measurements for your stream. Model development requires a suitable reference basis, application-relevant samples, coverage of expected concentrations including the extremes, and evidence that the model addresses the intended process conditions.
This branch is appropriate when the process contains multiple potential absorbers, when direct interpretation is uncertain, or when the intended use includes process or installation variation that must be represented. State the model scope, reference method, expected range, process conditions, validation samples, residual monitoring, and operating checks in the request. The supported boundary is that a model should not be treated as ready for use merely because it fits the initial calibration samples.
Request a technical discussion when you cannot yet map the stream to a configuration, when the sample route is uncertain, or when pressure, temperature, solids, reference data, or validation conditions remain undefined.
Bring the supplier-test brief to that discussion. A useful technical question includes the target result, sample path, concentration range, extremes, operating conditions, reference method, available samples, installation context, and validation questions. End the discussion with a defined evaluation scope rather than a general assumption that the instrument will fit.
Treat an instrument loan as a separate option that requires confirmation from Paeonia Innovations. If a loan is being considered, ask whether that option is available for your situation and what sample, configuration, model, validation, documentation, and operating conditions would apply. Availability and terms are not established by the general measurement principles in this guide.
Use an instrument loan only within a defined evaluation question. The evaluation should still identify the target measurement, representative stream, reference method, concentration range, installation context, and evidence required before you interpret the results as relevant to the intended process.
Prepare the technical discussion around your process evidence
The decision is ready for a technical discussion when you can state what must be measured, where the sample will come from, which conditions must be represented, and how reference data will support the model. Mid-infrared spectroscopy can support real-time monitoring in complex fermentation fluids, but the concentration result remains bounded by the reference method, calibration coverage, validation evidence, and intended process conditions.
A qualified technical or process professional must determine individual suitability for your application, including any process, regulatory, or validation requirements. Bring the completed brief so the proposed configuration and evaluation can be discussed against your actual process evidence.
What should you confirm before choosing a configuration?
Can a mid-infrared spectrometer measure fermentation concentrations directly?
It records an absorption spectrum; it does not automatically identify a substrate or product concentration. Concentration estimates require paired reference measurements, a suitable multivariate or chemometric model, and calibration and validation data that represent the intended stream and operating conditions.
What information should be included in a supplier test brief for bioprocess monitoring?
State the target measurement, expected concentration range and extremes, stream composition, phases and solids, temperature, pressure, sampling location, proposed sample path, reference method, available samples, installation context, and validation requirements. This lets the technical assessment address process fit rather than instrument specifications alone.
When should a bioprocess team request an instrument loan?
Consider a loan when representative on-site or pilot measurements are needed to examine sampling, matrix effects, installation constraints, or model transfer before committing to a permanent configuration. Loan availability, eligibility, duration, and terms must be confirmed with Paeonia Innovations, and the evaluation should still have a defined target, reference method, sample set, and validation scope.
Can a chemometric model developed in the laboratory be used in a pressurized installation?
Paeonia Innovations states that the OrionIR® High-Pressure Version uses the same measurement principle as its Lab version, supporting transfer of models developed during evaluation. However, empirical models can be affected by process, raw-material, sampling, and instrument variation, so application-relevant calibration and validation are still needed before relying on the model under pressure.