Literature Review · Peer-Reviewed Sources Only

Optek-Danulat ASD-N Biomass Probe: Performance Review from Four Peer-Reviewed Studies

Published 29 July 2026 · 13 min read · 4 citations
Optek-Danulat ASD-N NIR-absorption biomass probe — in-bioreactor schematic Stirred-tank fermenter LED det. ASD-N NIR absorption Hybrid-LED source at 840–910 nm; transmitted light → silicon photodiode Total biomass • Live cells attenuate the beam • So do dead cells / fragments → not viable-only Path length sets range OPL 1, 5, 10 or 20 mm 10 mm → linear to ~90 g/L DCW; degrades >90 g/L (Grigs 2021, P. pastoris) Optek-Danulat ASD-N NIR-absorption biomass probe
Figure 1: Optek ASD-N probe inserted through the headplate of a stirred-tank fermenter. A hybrid LED emits at 840–910 nm across a fixed optical path (1, 5, 10 or 20 mm depending on target cell density) to a silicon photodiode. Attenuation of the transmitted beam is proportional to total suspended biomass — live and dead cells both absorb, so the reading is total, not viable, biomass. Path length is the primary tuning parameter: 10 mm gives an ~90 g/L DCW linear ceiling in P. pastoris fed-batch [1].
Literature Verdict

Across four peer-reviewed studies — a P. pastoris fed-batch soft-sensor study up to 135 g/L DCW, a microalgal photobioreactor optical-methods benchmark, a 2025 hybrid-modelling and Model Predictive Control build on the same P. pastoris platform, and a classical NIR-vs-dielectric-spectroscopy comparison — the Optek-Danulat ASD-N range emerges as the most-cited transmitted-light NIR-absorption biomass probe for fermentation and microalgal bioprocesses. Grigs et al. 2021 reported an 8% normalised RMSE in P. pastoris DCW across the whole 0–135 g/L range and about 5 g/L RMSE below 50 g/L [1]. Havlik et al. 2022 measured R² = 0.97–0.99 between ASD19-N turbidity and off-line spectrophotometer OD in microalgal cultivation, matching the accuracy of the reference spectrophotometer itself [2]. Bolmanis et al. 2025 embedded the same ASD19-EB-01 signal into a hybrid deep-learning model for Model Predictive Control of the P. pastoris process [3]. The recurring caveats are Beer–Lambert saturation above about 90 g/L DCW with a 10 mm optical path, the total-vs-viable biomass difference against Aber/Hamilton capacitance probes [4], and unreliability in unclear media or cultures containing adsorbents or a solid growth matrix — the boundary Kiviharju et al. 2007 identified for NIR light absorption compared with dielectric spectroscopy.

Optek ASD-N range at a glance

The ASD-N family from Optek-Danulat GmbH (Essen, Germany, part of TT Electronics) is a set of single-channel NIR-absorption biomass probes for insertion through the headplate or side port of a fermenter or bioreactor. The three insertion probes share the same measurement principle — light absorption at 840–910 nm across a fixed optical path length between a hybrid LED source and a hermetically sealed silicon photodiode — and differ in probe body diameter, sanitary connection, and sterilisation compatibility. The spec table below is compiled from the ASD12-N datasheet and the ASD25-N datasheet; field performance evidence comes from the peer-reviewed studies cited throughout this review.

SpecificationASD12-N (pilot / R&D)ASD19-N (mid-scale)ASD25-N (production, sanitary)
Measurement principleSingle-channel light absorption; hybrid LED source at 840–910 nm; hermetically sealed silicon photodiode; seal-less sapphire window
Optical path length1, 5, 10 or 20 mm (choose shorter for denser bacterial / yeast cultures, longer for mammalian / algal)
Insertion depth110, 215 or 315 mm + OPLSimilar to ASD12 body, sized for larger vessels35 mm + OPL with 60 mm port length (30 mm extension available)
Process connection12 mm diameter, PG 13.5 thread19 mm diameter, M26 × 1 threadG 1¼ in. ISO 228/1 thread, similar-to-Ingold port, 3-A sanitary
SterilisationAutoclavable without cable; max 4 bar / 135°C; up to 60 min/dayAutoclavable (same construction as ASD12)CIP/SIP-optimised (3-A); autoclaving NOT possible
Process temperature5–50°C permanent (135°C for autoclave)Similar to ASD125–65°C permanent; 5–135°C peak (SIP)
Process pressurePressure-free (±0.5 bar)Pressure-free (as ASD12)0–10 bar (0–145 psi)
ConverterC4000 or C8000 platform; the hybrid-LED design allows up to 4 ASD25 probes per C4000
Output4–20 mA analogue; digital via converter (RS-232 / Ethernet); user-programmable measurement units (CU / OD / g/L after calibration)
Typical usePilot fermenters, single-use bag bioreactors, R&DMid-scale fermenters and reusable bioreactorsGMP / production fermenters and cGMP-validated cell culture
Indicative capital costApproximately €7,000–€15,000 per channel including converter (industry-typical for a stainless in-line NIR probe; contact Optek or your local distributor via the product page for a quote)

Spec values are taken directly from the Optek ASD12-N and ASD25-N product pages and the C4000/C8000 converter product information brochure. The INFORS HT integration page confirms the wavelength range and OPL family. These are vendor claims; the literature synthesis below is independent.

What the peer-reviewed literature says

Four studies from 2007 to 2025 provide directly quantitative or comparative evidence on the ASD-N range: two Latvia–Lithuania group papers on Pichia pastoris fed-batch that name the specific model (Grigs 2021 and Bolmanis 2025), one German-Colorado optical methods benchmark on microalgal cultivation that names the ASD19-N model (Havlik 2022), and a Finnish comparison paper from 2007 that benchmarked the NIR-absorption approach itself against dielectric spectroscopy in bioreactor microorganism cultivations (Kiviharju). The 2007 paper does not name Optek in its abstract, but it defines the failure boundary for any transmitted-light NIR absorption probe of the ASD-N type and is included here as the classical methodological reference the field still cites for that boundary. Together the four papers cover microbial and photobiological modes, dry-cell-weight ranges from below 1 g/L to 135 g/L, and both traditional linear regression and modern hybrid-deep-learning use of the probe signal.

Grigs, Bolmanis, Galvanauskas (2021) is the quantitative benchmark of record for the ASD19-EB-01 in high-cell-density yeast fed-batch [1]. The authors used the probe with a 10 mm optical path in a 5 L stirred-tank bioreactor across recombinant P. pastoris GS115 processes producing HBcAg (Mut+) and HBsAg (MutS), with glycerol and methanol substrates. The turbidity signal was correlated to off-line dry cell weight via an exponential calibration; the resulting biomass estimate held to an 8% normalised RMSE across the whole 0–135 g/L DCW envelope. Below 50 g/L, the RMSE was approximately 5.0 g/L; between 50 and 135 g/L, RMSE climbed to about 12.3 g/L. The authors identified 90 g/L DCW as the practical linear ceiling under constant mixing and aeration and explicitly recommended combining the ASD signal with soft-sensors (oxygen uptake, base addition) to extend usable range at higher cell density. This is the strongest quantitative-accuracy citation for the ASD-N range in the literature reviewed here, and it establishes the Beer–Lambert saturation ceiling that constrains any transmitted-light NIR probe.

Havlik, Beutel, Scheper, Reardon (2022) provides the tightest published R² number for the ASD19-N against a reference off-line method [2]. The paper is a review of on-line optical methods for microalgal bioprocesses that also reports the authors' own comparison of an RGB image sensor, an off-line spectrophotometer, and the commercial Optek ASD19-N turbidity probe in a closed flat-panel photobioreactor. The three methods gave comparable accuracy and precision, with R² = 0.97–0.99 across the tested range. The paper's practical contribution is confirming that even in the challenging photobioreactor environment — strong background chlorophyll pigmentation and gas-lift bubble noise — the ASD19-N's transmitted-light NIR geometry stayed linear against the spectrophotometer, in line with Optek's design claim that transmitted-light NIR at 840–910 nm is less susceptible to bubble artefacts and window coating than backscatter at visible wavelengths.

Bolmanis, Grigs, Galvanauskas (2025) built directly on the same platform, extending the earlier 2021 work with a hybrid deep-learning model and Model Predictive Control [3]. The same in-situ ASD19-EB-01 probe supplied the real-time biomass signal that anchored the hybrid model; the turbidity trace was correlated to biomass via an exponential calibration equation that the authors' MPC then used to close the control loop. The paper does not itself re-report the linearity numbers from the 2021 study — it treats the ASD probe as a validated primary sensor and focuses on the modelling and control layer above it — but its methodological choice is important evidence that the ASD probe reading is stable enough to serve as an anchoring signal for closed-loop MPC in cGMP-relevant P. pastoris production. It is the most recent (2025) use of the probe in the peer-reviewed literature reviewed here.

Kiviharju, Salonen, Moilanen, Meskanen, Leisola, Eerikäinen (2007) is the classical NIR-vs-dielectric comparison of the field [4]. The authors evaluated an on-line NIR light-absorption probe and an on-line dielectric-spectroscopy probe in bioreactor cultivations of several microorganisms and concluded that the NIR absorption method was — at the time — more developed and reliable for on-line in-situ biomass estimation than dielectric spectroscopy, with a critical explicit exception: NIR absorption is of no significant use when the cultivation medium is not clear, and especially in processes using adsorbents or a solid matrix for the microorganism to grow on. This is the reference boundary condition every subsequent transmitted-light NIR probe study — including the ASD-N literature above — has had to work within. The paper's abstract does not name the specific commercial probe model (Optek is not stated in the freely-visible portion), and we cite it here as methodological rather than product-attributive evidence — for the general NIR-vs-dielectric trade-off statement, not for a specific Optek performance number.

Performance data from cited studies

Study Conditions Accuracy / correlation Response / drift Conclusion
Grigs 2021 [1] P. pastoris GS115 HBcAg (Mut+) / HBsAg (MutS) fed-batch, 5 L stirred-tank, glycerol + methanol substrates; Optek-Danulat ASD19-EB-01 with 10 mm OPL NRMSE 8% across 0–135 g/L DCW; RMSE ~5.0 g/L below 50 g/L; RMSE ~12.3 g/L between 50 and 135 g/L; exponential calibration to off-line DCW Acceptable up to ~90 g/L DCW under constant mixing / aeration; degradation is Beer–Lambert saturation, not probe drift ASD19 signal reliable primary sensor at moderate cell density; needs soft-sensor coupling above ~90 g/L DCW
Havlik 2022 [2] Microalgal cultivation in a closed flat-panel photobioreactor; three-way comparison of RGB image sensor, off-line spectrophotometer, and commercial Optek ASD19-N NIR probe R² = 0.97–0.99 for ASD19-N vs off-line spectrophotometer OD; accuracy and precision comparable to spectrophotometer and RGB sensor Stable across the tested cultivation; no calibration drift reported over the run Transmitted-light NIR remains linear against reference OD in photobioreactor conditions, in line with the vendor claim of bubble and coating tolerance
Bolmanis 2025 [3] Recombinant P. pastoris fed-batch (same platform as Grigs 2021); Optek-Danulat ASD19-EB-01 anchoring a hybrid deep-learning + Model Predictive Control loop ASD signal anchored a closed-loop MPC via an exponential turbidity–biomass calibration; specific accuracy numbers inherited from the 2021 platform paper Probe used as validated primary sensor for real-time MPC; no drift correction event reported ASD-N reading is stable enough to sit at the base of a hybrid-model MPC in cGMP-relevant P. pastoris production
Kiviharju 2007 [4] Bioreactor cultivations of several microorganisms; on-line NIR light-absorption probe compared with an on-line dielectric-spectroscopy probe (comparative methodology paper) Not attributed to the ASD-N range specifically; general NIR-absorption vs dielectric comparison paper NIR method identified as more developed and reliable than dielectric spectroscopy in clear media at the time of publication NIR absorption fails when the cultivation medium is not clear or in processes using adsorbents / solid matrix — the boundary the ASD-N literature above operates within

Every row is a separate peer-reviewed publication; see References section for full citations. Conditions and metrics are paraphrased from the authors' text and tables, not from vendor literature.

Limitations and failure modes reported

Across the reviewed studies, and reading carefully for what the authors do not claim, the following limitations and failure modes recur. Each bullet is tagged with the citations that describe it.

When the literature recommends Optek ASD-N

Recommended for

  • High-cell-density microbial fed-batch (P. pastoris, E. coli, yeast) where total biomass tracking below ~90 g/L DCW is the meaningful control variable and 10 mm optical path is a fit [1]
  • Microalgal photobioreactors and clear-media processes where transmitted-light NIR at 840–910 nm avoids visible-spectrum interference from chlorophyll pigmentation [2]
  • Hybrid modelling and Model Predictive Control installations that need a stable, calibrated real-time biomass signal to close the control loop [3]
  • Bubbly, high-agitation processes where backscatter turbidity probes struggle — transmitted-light geometry with a sapphire seal-less window minimises bubble noise and window coating [2]

Caveats / not recommended for

  • Very-high-density microbial fed-batch above ~90 g/L DCW with a single-probe expectation — the linear ceiling requires soft-sensor coupling or a shorter optical path [1]
  • Processes where viable cell density — not total biomass — is the KPI (late-stage CHO fed-batch, apoptosis-sensitive perfusion) — use a capacitance probe instead [4]
  • Unclear or coloured media (lignocellulose hydrolysates, melanin producers, adsorbent-containing systems, solid-matrix immobilised-cell processes) — the classical NIR-absorption failure boundary [4]
  • cGMP CHO fed-batch decisions that require published peer-reviewed CHO validation of the ASD-N — that evidence is missing from the four papers reviewed here (present only in vendor literature) [1]

Use cases documented in the literature

Specific deployments reported in the cited studies. Each card corresponds to a real published bioprocess use case.

High-density yeast fed-batch
P. pastoris 0–135 g/L DCW soft-sensor

ASD19-EB-01 with 10 mm OPL benchmarked against off-line DCW in HBcAg / HBsAg P. pastoris fed-batch; 8% NRMSE across the whole range; ~5 g/L RMSE below 50 g/L.

[1]
Microalgal photobioreactor
ASD19-N vs RGB sensor vs off-line spectrophotometer

Three-way accuracy comparison in a closed flat-panel photobioreactor; ASD19-N reached R² = 0.97–0.99 vs off-line OD; matches spectrophotometer accuracy in bubbly, pigmented broth.

[2]
Hybrid MPC / PAT
P. pastoris Model Predictive Control

ASD19-EB-01 anchored a hybrid deep-learning model and Model Predictive Control loop in P. pastoris fed-batch; validated primary sensor for closed-loop biomass control.

[3]
Methodology benchmark
NIR absorption vs dielectric spectroscopy

Classical comparison of transmitted-light NIR against dielectric spectroscopy across multiple microorganisms; identified the "unclear media / adsorbent / solid matrix" failure boundary that later ASD-N deployments respect.

[4]

Comparing Optek ASD-N against alternatives?

The Sensor Selection Tool takes 6 questions about your scale, modality, vessel, and budget and returns ranked sensor recommendations — capacitance (Aber FUTURA, Hamilton INCYTE), scattered-light turbidity (Hamilton Dencytee), non-invasive shake-flask sensors, and stainless single-use options included.

Open the Sensor Selection Tool

User reviews from bioprocess engineers

Real-world experience from engineers who deployed the Optek-Danulat ASD-N NIR biomass probe. All reviews are moderated before publishing. Share your own below — 2 minutes, anonymous option available.

Frequently asked questions

What does the Optek-Danulat ASD-N biomass probe actually measure?
The Optek ASD-N series measures the near-infrared (NIR) light absorption of the broth as light passes across a fixed optical path length between a hybrid-LED source and a silicon photodiode. At 840–910 nm the medium itself is largely transparent, so attenuation is dominated by scattering-and-absorption from suspended cells — making the signal proportional to total biomass concentration (live plus dead cells, plus cell fragments). Unlike Aber-style capacitance probes, the ASD signal does not distinguish viable from non-viable biomass; unlike backscatter turbidimeters, it uses transmitted (not reflected) light, which the vendor states makes the reading less sensitive to bubble noise and window fouling. Optical path length is chosen to match cell density — 1 or 5 mm for dense bacterial or yeast cultures, 10 or 20 mm for lower-density mammalian or microalgal work.
How accurate is the Optek ASD19-N against offline dry cell weight?
Grigs et al. 2021 benchmarked an ASD19-EB-01 with a 10 mm optical path against offline dry cell weight in P. pastoris fed-batch cultivations up to 135 g/L DCW [1]. Below 50 g/L DCW the RMSE was approximately 5.0 g/L; across the whole 0–135 g/L range the normalised RMSE was 8%. In microalgal photobioreactors, Havlik et al. 2022 report R² = 0.97–0.99 between ASD19-N turbidity and off-line spectrophotometer optical density — the same correlation strength as an RGB image sensor and the reference spectrophotometer themselves [2]. In both regimes the linear range is bounded by the optical path length: too long and you saturate absorbance at moderate density; too short and low-density noise dominates.
When should I choose ASD12-N, ASD19-N, or ASD25-N?
ASD12-N is a 12 mm-diameter probe with a PG 13.5 thread that fits any bioreactor headplate port — the pilot / R&D probe, autoclavable (max 4 bar, 135°C for up to 60 min/day) and available in three insertion lengths (110, 215, 315 mm + OPL). ASD19-N is the mid-scale sibling in a 19 mm / M26×1 port, typically used in larger fermenters or bioreactors where the 12 mm port is not enough mechanical rigidity. ASD25-N is the production-scale, sanitary probe: G 1¼ in. ISO 228/1 thread, similar-to-Ingold port, 3-A sanitary CIP/SIP-optimised, rated 5–135°C (peak) and 0–10 bar, but NOT autoclavable. Below the ASD-N range, Optek also lists the AF16-N / TF16-N flow-cell turbidimeters for external loop or downstream monitoring rather than in-vessel insertion — see the Optek products catalogue.
Does the Optek ASD-N work in single-use bioreactors?
The ASD12-N is marketed for pilot-scale fermenters and single-use bioreactor bags via aseptic connectors on the vessel head. The probe body itself is stainless steel with a sapphire window and can be autoclaved separately then aseptically inserted through a PG 13.5 port on a compatible bag. However, unlike the true single-use variants (Aber FUTURA NEO or Hamilton Incyte SU), the ASD probe is a reusable insertion probe that is aseptically transferred between bags — not a pre-sterilised, gamma-irradiated, disposable sensor. For a comprehensive treatment of these architectures, see the single-use sensor review on this site.
What is the reported linear range and where does accuracy degrade?
Grigs et al. 2021 reported that the ASD19-EB-01 with a 10 mm optical path gave acceptable performance up to approximately 90 g/L DCW under constant mixing and aeration in P. pastoris fed-batch; from 50 to 135 g/L DCW the RMSE climbed from 5.0 g/L to approximately 12.3 g/L [1]. This degradation is the classical Beer–Lambert saturation effect — at high cell density, absorbance approaches the detector's dynamic-range ceiling and the transmitted-light signal becomes non-linear in biomass. The paper's recommendation for higher-density work is to combine the ASD probe with a soft-sensor (oxygen uptake rate, base addition) that provides orthogonal biomass information the NIR probe alone cannot resolve.
How does NIR absorption compare to capacitance and scattered-light biomass probes?
Kiviharju et al. 2007 directly compared NIR light-absorption and dielectric spectroscopy for on-line biomass measurement in bioreactor cultivations of several microorganisms [4]. The authors concluded the NIR absorption method was, at the time, more developed and reliable for on-line in-situ biomass estimation than dielectric spectroscopy — but with an explicit failure mode: NIR absorption becomes unreliable when the medium is not clear or when adsorbents or a solid matrix are present. Capacitance is the opposite trade-off (viable-only signal, higher robustness to non-cell background, but weaker response in dilute cultures and cell-state-sensitive in the decline phase). Backscatter turbidity probes trade linearity at high density for a reflected-light geometry that avoids Beer–Lambert saturation, but the vendor and Kiviharju agree they are more sensitive to bubble noise and window coating than the ASD transmitted-light design. See the capacitance vs optical biomass sensor comparison for the full trade-off matrix.
Is the Optek ASD-N autoclavable, CIP/SIP compatible, or both?
It depends on the model. The ASD12-N is autoclavable (without cable) at a maximum of 4 bar and 135°C for up to 60 minutes per day. The ASD19-N shares the same seal-less sapphire-window construction and is autoclave-compatible. The ASD25-N is the production probe: 3-A sanitary and CIP/SIP-optimised, with a permanent process temperature envelope of 5–65°C and a 5–135°C peak (for SIP), but the vendor datasheet explicitly states autoclaving is NOT possible for the ASD25-N. If your process needs autoclave sterilisation, choose the ASD12-N or ASD19-N; if you need in-place CIP/SIP with a sanitary connection, choose the ASD25-N. All models use the same 840–910 nm hybrid-LED source and the same C4000 / C8000 converter platform.
What are the reported limitations of the Optek ASD-N in bioreactor use?
Four limitations recur across the reviewed literature. First, absorbance saturates at very high cell density — Grigs 2021 reports acceptable P. pastoris performance up to about 90 g/L DCW but degraded RMSE from 5 to 12 g/L in the 50–135 g/L range [1]. Second, the probe reports total (not viable) biomass, so late-culture divergence between offline VCD and the ASD reading is expected in fed-batch mammalian and microbial processes where dead-cell fraction rises. Third, Kiviharju 2007 noted that NIR absorption becomes unreliable in unclear media or adsorbent-containing cultures [4]. Fourth, the Havlik 2022 review and Bolmanis 2025 both use soft-sensor and hybrid-modelling strategies that combine the ASD signal with orthogonal on-line measurements — a practical acknowledgement that ASD alone is not sufficient for tight biomass control at industrial scale [2][3].

References

  1. Grigs, O., Bolmanis, E., Galvanauskas, V. (2021). Application of In-Situ and Soft-Sensors for Estimation of Recombinant P. pastoris GS115 Biomass Concentration: A Case Analysis of HBcAg (Mut+) and HBsAg (MutS) Production Processes under Varying Conditions. Sensors, 21(4):1268. DOI: 10.3390/s21041268.
  2. Havlik, I., Beutel, S., Scheper, T., Reardon, K. F. (2022). On-Line Monitoring of Biological Parameters in Microalgal Bioprocesses Using Optical Methods. Energies, 15(3):875. DOI: 10.3390/en15030875.
  3. Bolmanis, E., Grigs, O., Galvanauskas, V. (2025). Leveraging Historical Process Data for Recombinant P. pastoris Fermentation Hybrid Deep Modeling and Model Predictive Control Development. Fermentation, 11(7):411. DOI: 10.3390/fermentation11070411.
  4. Kiviharju, K., Salonen, K., Moilanen, U., Meskanen, E., Leisola, M., Eerikäinen, T. (2007). On-line biomass measurements in bioreactor cultivations: comparison study of two on-line probes. Journal of Industrial Microbiology & Biotechnology, 34(8):561–566. DOI: 10.1007/s10295-007-0233-5.