Inline NMMO solvent concentration in lyocell fibre production

Man-made fibres · Lyocell · Solvent recovery

NMMO in lyocell spin baths, evaporators and recycled solvent: refractive index is linear in NMMO, about 0.0018 per wt% at 60 °C; grade by temperature, material by thermal hazard.

Applicable industries
Inline NMMO solvent concentration in lyocell fibre production

PISONICS

PS7110 Series

Inline Process Refractometer

Critical-angle refractometry · sapphire prism · temperature compensation

Inline NMMO Solvent Concentration in Lyocell Fibre Production

—— Spin bath · solvent recovery evaporation · recycled solvent ——

Key measurements: NMMO wt% / refractive index nD / temperature °C

Lyocell Fibre: Inline NMMO Solvent Concentration

1. Why the solvent concentration matters

The lyocell process dissolves cellulose directly in aqueous N-methylmorpholine N-oxide (NMMO): the spinning dope is roughly 76 : 10 : 14 NMMO, water and cellulose; the filaments enter the spin bath, the NMMO is washed out, purified and concentrated again by multi-effect evaporation for reuse. Commercial NMMO is usually a 50 wt% aqueous solution.

Almost all of the solvent circulates. A published patent estimates about 90 t of water evaporated per tonne of fibre; major producers report solvent recovery of about 99.7 % or higher. Make-up water to the spin bath, the discharge of each evaporator effect and the recycled solvent are all controlled by concentration measurement. According to the China Chemical Fibers Association, Chinese lyocell output passed 400,000 t in 2024, and China holds over 60 % of world capacity.

2. Refractive index is linear in NMMO concentration

A published patent gives two points at 60 °C: 25 % aqueous NMMO at nD = 1.3676 and 78 % at 1.4624, linear in between. From these, at 60 °C one wt% of NMMO moves the refractive index by about 0.0018.

SystemTemp.Refractive index nDSource
NMMO / water 25 / 7560 °C1.3676Published patent
NMMO / water 78 / 2260 °C1.4624 (process tolerance 1.4620–1.4628)Published patent
NMMO 50 % (commercial solution)20 °C1.4201Reagent product data
Dope: cellulose / water / NMMO = 15 / 9 / 7660 °C1.4895 ± 0.0005Published patent

The wavelength is not stated in the sources; refractive index changes with temperature, so inline measurement must be temperature-compensated.

Two things follow. The sensitivity is sufficient: the 78 % solvent tolerance of ±0.0004 is about ±0.2 wt%, and the PS7110's ±0.0001 in refractive index converts to about 0.06 wt% on that slope (a unit conversion only). And the scale has to be built for NMMO: the PS7110 leaves the factory with a %Brix scale, so a wt% scale for NMMO, with its temperature compensation, is built on the real liquid.

3. Measuring points and grades

PointTypical concentration / temperature (published)What the inline value is forGrade
Spin bathCommonly 20–25 % (processes differ); bath at about 5–30 °CMake-up water controlA
Dilute recovery liquorUsually below 10 %Recovery loadA
Evaporator effect dischargeConcentrated stage by stage to 50–55 %; effects from about 50 °C to 120 °CEvaporation end pointB below 70 °C, C at 70–100 °C; effects above 100 °C are out of range
Recycled solventAbout 50–55 %, or concentrated further to 70–80 %Check before reuseB / C (concentrated solvent is close to the top of grade A)
Spinning dopeAbout 76 % NMMO with cellulose; often 90–120 °CDegree of dissolution, for a fixed recipeGrade C can be assessed up to 100 °C; see section 4
GradePrismProcess temp.RI rangeFactory scaleFits
PS7110-AHigh-strength optical glass0–60 °C1.33299–1.465100–70 %BrixSpin bath, dilute liquor
PS7110-BSapphire0–70 °C1.33299–1.517820–90 %BrixConcentrated solvent below 70 °C
PS7110-CSapphire−10–100 °C1.33299–1.570410–100 %BrixEvaporator discharge and recycled solvent at 70–100 °C

4. Three things to settle first

  • Thermal safety and wetted materials. NMMO is thermally unstable: published work places accelerated degradation above 120 °C and possible runaway above 150 °C, and iron, copper and their salts lower the decomposition temperature markedly. A published patent holds that wetted surfaces should not contain copper, molybdenum, tungsten, cobalt and similar elements; the PS7110's standard 316L contains molybdenum (about 2–3 %), and so does Hastelloy. Where that restriction applies, grades B / C can be ordered with titanium wetted parts; confirm the final material with the process licensor. The instrument does not heat the medium.
  • Crystallization. Concentrated NMMO crystallizes on cooling (the monohydrate melts at about 74 °C); crystals on the prism falsify the reading and can block sample lines. Trace and lag concentrated-solvent points and sample lines, and keep the point above the crystallization temperature.
  • Impurities and the ternary dope. Recovered liquor carries degradation products such as NMM and morpholine (NMM at about 0.1–0.3 % in recovered concentrate), and they contribute to the refractive index too; colour from the stabilizer does not affect critical-angle refractometry. Compare the inline reading with titration or HPLC regularly. The dope is a ternary NMMO / water / cellulose system: refractive index alone shows the degree of dissolution for a fixed recipe, not the three proportions.

5. Installation

  • Never mount the detection face pointing down — deposits build up, and a deposit does not raise a fault; it shows only as a slow one-way drift
  • On a vertical line the flow must be upward — a downcomer drains when the pump stops, the prism is no longer wetted and the reading means nothing
  • On a horizontal line insert from below (face up) or from the side (face vertical); inserting from the top is not allowed
  • Downstream of a make-up water or solvent inlet, leave 10 D upstream and 5 D downstream (at least 5 D and 3 D) so the liquid is mixed before it is measured
  • Prefer a bypass with valves at both ends, so the instrument can be isolated without stopping the main line
  • Where the solvent area is a hazardous area: the PS7110 is intrinsically safe, Ex ia IIC T6 Ga; for hot media the temperature class is set by the maximum medium temperature

PS7110 key specifications

ItemSpecification (data sheet Rev. B)
Measured valueRefractive index nD; converted to concentration by a scale (%Brix, or a wt% scale built for the medium)
AccuracyRI ±0.0001; Brix ±0.1 %; temperature ±0.5 °C
ResolutionRI 0.00001; Brix 0.01 %; temperature 0.1 °C
Measuring interval2–60 s, settable
Pressure ratingDetection surface ≤ 1.5 MPa
Cleaning temperature0–120 °C, CIP / SIP capable
Wetted materialsSS316L + prism; Hastelloy, titanium or tantalum optional on grades B / C
Process connectionClamp (Ø77.5 mm ferrule); custom tee / cross, cross in DN25 / 50 / 65 / 80
Outputs4–20 mA; RS485 (Modbus RTU)
Power / ambientDC 24 V; −10 to 70 °C
Hazardous areaEx ia IIC T6 Ga (intrinsically safe); for hot media the temperature class is set by the maximum medium temperature

Before you order

  • NMMO concentration range and temperature at each point, including the maximum during start-up, shutdown and cleaning.
  • Whether the liquid carries cellulose, gel particles or precipitate.
  • Line size, pressure and flow condition (running full or not).
  • Current lab method (titration, HPLC or bench refractometer) and how often it is compared.
  • Wetted-material restrictions (copper, molybdenum and the like), hazardous area, required outputs.

What this page is: a method and selection note, not a case study. The PS7110 has no publishable operating data on this medium yet, so there are no unit counts, saving percentages or payback periods here — we do not invent those. The numbers on the page are of three kinds only: clauses of standards, published property data with the estimates derived from them (marked as estimates or calculations), and the instrument's own specification.

Conclusion

PISONICS covers six principle families (ultrasonic, tuning fork, Coriolis, optical, microwave, differential pressure) for inline density and concentration; the first step of any selection is to ask what quantity the process actually needs, and in which range. Binary NMMO–water solutions are a textbook use of refractometry, in the same family as DMF and NMP in solvent recovery. Further reading: optical concentration meters: refractometry or spectroscopy, DMF / DMAc / NMP solvent recovery, optical concentration meters.

FAQ

Do impurities in recovered NMMO bias a refractometer?

They can, so compare regularly. Recovered liquor carries degradation products such as NMM and morpholine (NMM at about 0.1–0.3 % in recovered concentrate), and anything dissolved contributes to the refractive index; colour from the stabilizer does not affect critical-angle refractometry.

Pair the inline reading with titration or HPLC results at intervals, and correct the scale from the comparison.

Why heat-trace a concentrated NMMO measuring point?

Concentrated NMMO crystallizes on cooling (the monohydrate melts at about 74 °C); crystals on the prism falsify the reading and can block sample lines. Trace and lag concentrated-solvent points and sample lines so the point stays above the crystallization temperature.

Keep within the process limit of the PS7110 grade — A 60 °C, B 70 °C, C 100 °C — and remember that NMMO itself degrades faster above 120 °C, so the tracing should not run hot.

Can the PS7110 measure the spinning dope directly?

Published patents use refractive index to judge how far the dope has dissolved (mostly on samples; 1.4895 ± 0.0005 at 60 °C as the pass band). But the dope is a ternary NMMO / water / cellulose system, and refractive index alone shows dissolution only for a fixed recipe; the dope also often runs at 90–120 °C, while the PS7110 is rated to 100 °C (grade C).

Only dope points up to 100 °C can be assessed; hotter points are outside the PS7110's range, and because the dope is very viscous the point must also keep fresh liquid flowing over the detection face.

Does the PS7110 prism need cleaning, and how often should it be checked?

A clean prism face is the single precondition for a trustworthy reading. Coating, scale, crystals or abrasion change the reading directly, and the instrument does NOT raise a fault when they do — it shows up as a slow one-way drift on an unusually smooth curve. So the cleaning strategy belongs in the design, not in a response to a reading that has already drifted. The prism material depends on the variant: high-strength optical glass on the A, sapphire on the B and C. For media that coat or crystallise, the PS7110-CL1 ultrasonic cleaning unit keeps the face clean continuously and the PS7110-CL2 high-pressure flushing unit flushes it periodically; sanitary processes can use CIP and SIP directly, with a cleaning temperature range of −30 to 120 °C. Where the medium can carry grit or abrasive particles, filter them out upstream or the detection window will be worn away. Putting "look at the face" into the routine inspection is worth more than any coefficient correction applied afterwards; the sensor status diagnostics help, but they do not replace looking.