Osmolar

How osmometers work

Freezing point depression vs vapor pressure osmometry: principles, calibration and error sources.

An osmometer measures the osmolar concentration of a solution, meaning the number of dissolved particles. Clinical laboratories normally report osmolality (mOsm/kg water) rather than osmolarity (mOsm/L), because a mass-based quantity does not change with temperature [1].

Dissolving a solute changes four colligative properties of water, each in proportion to the number of particles [1]:

  • it lowers the freezing point
  • it lowers the vapor pressure
  • it raises the boiling point
  • it raises the osmotic pressure

Commercial instruments use three of these: freezing point, vapor pressure and membrane (osmotic pressure) methods [1]. For the units, see osmolality vs osmolarity.

1. Freezing-point depression

Principle. One osmole of dissolved particles per kilogram of water lowers the freezing point by about 1.86 °C, which is the cryoscopic constant of water, K_f ≈ 1.86 °C·kg/mol [1, 2]:

ΔT_f = K_f × osmolality, so osmolality (Osm/kg) = ΔT_f ÷ 1.86

How the instrument works. A small sample is cooled below its freezing point (supercooled). Crystallization is then triggered, and the heat of fusion released brings the temperature up to a plateau at the true freezing point [2]. A thermistor in a bridge circuit reads that temperature [1]. The instrument compares the plateau with pure water and converts the difference to mOsm/kg [1]. Freezing-point depression is the method most clinical laboratories use, chosen for its accuracy and ease of use [1]. StatPearls gives 0–2,000 mOsm/kg as a common reportable range [1].

Worked example 1 (osmolality → freezing point). A serum sample of 290 mOsm/kg:

  1. 290 mOsm/kg = 0.290 Osm/kg
  2. ΔT_f = 1.86 × 0.290 = 0.539 °C
  3. The sample freezes at about −0.539 °C

Worked example 2 (freezing point → osmolality). The instrument reads ΔT_f = 0.558 °C:

  1. 0.558 ÷ 1.86 = 0.300 Osm/kg = 300 mOsm/kg

Worked example 3 (ideal vs real). The label value for 0.9% NaCl, 308 mOsm/L, is a calculated ideal osmolarity. Measured osmolality is about 286 mOsm/kg, because NaCl behaves non-ideally (osmotic coefficient ≈ 0.93) [3, 4].

  • Ideal value: 1.86 × 0.308 = 0.573 °C
  • Measured value: 1.86 × 0.286 = 0.532 °C

An osmometer reports the real particle activity, not the ideal count.

Interferences (StatPearls) [1]:

  • Particulate matter can cause premature crystallization. Centrifuging, or filtering urine, avoids this.
  • Air bubbles can prevent adequate freezing.
  • Samples that have evaporated, are non-aqueous, or are highly viscous may not freeze well.
  • Very high concentrations of ethanol, methanol, acetone, paraldehyde, trichloroethane or propylene glycol may not freeze reliably.
  • Carry-over from an uncleaned sample chamber skews the next result.

2. Vapor-pressure (dew-point) osmometry

Principle. Dissolved particles lower the vapor pressure of water. A vapor-pressure osmometer equilibrates a small sample in a closed chamber and measures the resulting temperature difference, which it converts to osmolality. StatPearls describes the voltage difference between 2 thermistors [1]; the VAPRO uses a thermocouple [5]. It works at room temperature, without freezing the sample [5].

Why it misses volatile alcohols. A volatile solute such as ethanol or methanol enters the vapor phase itself, so it does not lower the vapor pressure in the way a non-volatile solute does. The method therefore under-reads or ignores it:

  • Walker et al. (1986): a patient with methanol poisoning had a markedly elevated methanol level but no osmolal gap. The serum had been measured on a vapor-pressure osmometer, and “this technique may not detect volatile solutes such as alcohols”. In vitro testing confirmed the insensitivity to toxic ranges of methanol and ethanol. The authors concluded that an osmolal gap “should be sought only if freezing point osmometry is available” [6].
  • Draviam et al. (1984): the difference between freezing-point and vapor-pressure osmolality “is due to the presence of volatiles in the serum, because the volatiles are not measured by vapor pressure osmometry”. They multiplied that difference by 4.2 to estimate ethanol in mg/dL, and also reported a linear relationship for isopropanol and methanol added to serum [7].
  • StatPearls: vapor-pressure osmometers “cannot be used to measure the osmolality of solutions containing volatile solutes” and are less precise than freezing-point instruments [1].

Worked example 4 (Draviam difference method, arithmetic only). Freezing-point osmolality 330 mOsm/kg, vapor-pressure osmolality 306 mOsm/kg.

  1. Δosm = 330 − 306 = 24
  2. 24 × 4.2 ≈ 101 mg/dL, Draviam’s estimate of ethanol

The 4.2 is that laboratory’s empirical factor. It differs from the factors used to add ethanol to a calculated osmolality, which are ÷ 3.7 (Purssell; Osmolar default) and ÷ 4.6 (theoretical) [8, 9]. The difference method is not a substitute for a measured ethanol level.

The practical rule: an osmolar gap needs a freezing-point measured osmolality [6, 8]. See osmolar gap.

Where vapor pressure has an advantage. Formulation scientists sometimes prefer it for viscous or heat-sensitive samples:

  • In antisense-oligonucleotide formulations, vapor pressure was “generally more suitable” above about 150 mg/mL because of viscosity, and the two methods were comparable otherwise [10].
  • For parenteral excipients, freezing-point values tended to be similar to or higher than vapor-pressure values. The authors suggested vapor pressure for formulations with high excipient concentrations or thermosensitive excipients [11].

3. Membrane (colloid) osmometers and oncotic pressure

A membrane osmometer separates the sample from a reference solution by a semipermeable membrane and measures the pressure that develops as solvent moves across it [1]. Clinical colloid osmometers use membranes that pass water and small ions but retain proteins, so they measure only the colloid osmotic (oncotic) pressure (COP) of plasma, in mmHg. One commercial instrument’s datasheet gives a membrane cut-off of 10,000–30,000 Da and a Ringer’s-solution reference chamber [12]. Membrane osmometers are limited by the measurable range and by available membrane materials [1].

Normal values. Morissette (1977), using a pressure transducer and membrane system, reported that normal human plasma COP averages 25.4 mmHg. It was lower with age, in women and with bed rest. Values derived from serum protein matched measured COP in steady state but were unreliable in pathological sera [13].

Why freezing point cannot see oncotic pressure (our arithmetic). By van ‘t Hoff, π = c·R·T. At 37 °C (310.15 K), with R = 62.3637 L·mmHg·mol⁻¹·K⁻¹, R·T ≈ 19,342 mmHg·L/mol, so 1 mOsm/L ≈ 19.3 mmHg (RT/1000 = 19.3421 mmHg per mOsm [17]). A COP of 25.4 mmHg [13] therefore corresponds to only about 1.3 mOsm/L of colloid (25.4 ÷ 19.342 = 1.31; an ideal van ‘t Hoff estimate, since plasma proteins behave non-ideally), under half a percent of a ~290 mOsm/kg plasma osmolality. A freezing-point osmometer cannot resolve that. A membrane that retains only the colloid can.

4. Calibration and quality control

The figures below are general StatPearls guidance [1]. Laboratories follow their own policies and the instrument manual.

  • Calibration. Calibrate with manufacturer-supplied solutions of known osmolality. A full calibration at least every 6 months, and again after part replacement or service, or if quality-control results are repeatedly out of range.
  • Daily QC. Run commercially prepared controls at low, normal and high osmolality. Laboratories are typically required to run two controls at two concentrations daily or with each batch. Acceptable results are generally within ±2 SD of the mean, and deviations trigger corrective action.
  • External proficiency testing. Blinded samples are compared across laboratories to detect systematic bias.
  • Documentation and training. Keep control results, calibration records and corrective actions on file, and train staff on operation and troubleshooting.

In pharmaceutical quality control, USP general chapter 〈785〉 Osmolality and Osmolarity includes a freezing-point osmometry method and a calibration check with standard solutions; a proposed companion chapter 〈1785〉 covers the use of different osmometer types on the same sample [14] (public briefing; full text requires a subscription). Example 2 above shows the arithmetic link: a standard that depresses the freezing point by 0.558 °C corresponds to 300 mOsm/kg.

5. Sample handling

Acceptable specimens include serum, heparinized plasma, urine, tears and sweat. Requirements differ by instrument [1]. General points [1]:

  • Centrifuge before analysis to remove particles. Urine usually needs no preservative.
  • Stability:
    • unspun serum or plasma: 3 h
    • separated serum or plasma: 48 h (room temperature or 2–8 °C)
    • urine: 24 h at room temperature, 7 days refrigerated or frozen
  • Prevent evaporation and air bubbles, and clean the sampler and chamber between runs.
  • Volatile samples may need duplicate analysis.

For an osmolar gap, collect the measured osmolality and the chemistry (sodium, glucose, BUN or urea, ethanol) on the same draw.

6. Clinical and pharmaceutical uses

Clinical laboratories use osmometry for:

For stool, StatPearls describes a stool osmotic gap as measured stool osmolality − 2 × (stool Na + K) [1]. A gap < 50 mOsm/kg implies secretory diarrhea and > 75 osmotic diarrhea. A stool osmolality below 290 can mean water was added, and above 330 can mean poor storage or contamination with urine [1].

Pharmaceutical and biopharmaceutical laboratories measure osmolality in:

  • parenteral products (USP 〈785〉) [14]
  • oligonucleotide formulations, where measured osmolality was about 70% of theoretical (osmotic coefficient ≈ 0.7) and ~140 mg/mL was near 290 mOsm/kg [10]
  • high-concentration protein formulations, where measured values “consistently exceeded theoretical predictions” [15]

These examples show why calculated osmolarity does not replace measurement.

Instrument types on the market (named for illustration, not endorsement; specifications as stated by the manufacturer or trade press):

  • Freezing point: Advanced Instruments’ OsmoPRO, a 20-sample freezing-point instrument using 20 µL per sample [16]
  • Vapor pressure: ELITechGroup’s VAPRO 5600, specified at 10 µL and 20–3,200 mOsmol/kg [5]
  • Colloid osmometer: Gonotec’s OSMOMAT 050 [12]

Other manufacturers make comparable instruments.

Practice questions

Q1. A urine sample freezes at −0.600 °C. What is its osmolality?

Answer

0.600 ÷ 1.86 = 0.3226 Osm/kg ≈ 323 mOsm/kg [1].

Q2. A patient may have swallowed methanol. The only available osmometer is a vapor-pressure instrument, and the osmolar gap is 2. What can you conclude?

Answer

Very little. Vapor-pressure osmometry may not detect volatile alcohols, so the gap can be falsely small [6, 7]. Gap work requires a freezing-point measured osmolality. Even then, a normal gap does not exclude toxic alcohol ingestion. See osmolar gap.

Q3. Plasma COP falls from 25 to 15 mmHg. Roughly how many mOsm/L of colloid is that, and would a freezing-point osmometer show it?

Answer

10 mmHg ÷ 19.3 mmHg per mOsm/L [17] ≈ 0.5 mOsm/L (ideal estimate). That is far below the resolution needed against ~290 mOsm/kg, so a freezing-point result would barely change. COP needs a colloid (membrane) osmometer [12, 13].

Q4. Today’s normal-level control reads 3.1 SD above its mean. What does StatPearls describe?

Answer

Results outside the pre-defined limits (generally ±2 SD) trigger corrective and preventive action. Repeatedly out-of-range QC is also a reason to recalibrate [1]. Follow the laboratory’s own QC policy before reporting patient results.

Educational content, not medical advice or laboratory SOP. Follow your instrument’s instructions for use, your laboratory’s quality-control policy and applicable pharmacopeial requirements. See our disclaimer and how we check content.

Related: Osmolality vs osmolarity · Osmolar gap · Serum osmolality calculator · Urine osmolality · IV fluid osmolarity · Isotonic, hypotonic, hypertonic

References

  1. Larkins MC, Zubair M, Thombare A. Osmometer. In: StatPearls [Internet]. Treasure Island (FL): StatPearls Publishing. https://www.ncbi.nlm.nih.gov/books/NBK589659/
  2. Centers for Disease Control and Prevention, National Center for Health Statistics. Laboratory Procedure Manual: Osmolality (Urine), NHANES 2009–2010 (freezing point depression osmometry; test principle). https://wwwn.cdc.gov/nchs/data/nhanes/public/2009/labmethods/UCOSMO_F_met.pdf
  3. Williams E. The difference between actual and calculated osmolality of IV solutions should not be overlooked [letter]. Anesth Analg. 1999;89(6):1592. doi:10.1097/00000539-199912000-00072. https://doi.org/10.1097/00000539-199912000-00072
  4. Zander R, Ziegenfuß T, Sümpelmann R. Osmolality (mosmol/kg H2O) versus osmolarity (mosmol/L): applied physiology to improve patient safety. Eur J Med Res. 2025;30:1227. doi:10.1186/s40001-025-03652-7. https://doi.org/10.1186/s40001-025-03652-7
  5. ELITechGroup. VAPRO® 5600 Vapor Pressure Osmometer (product specifications). https://www.osmometers.com/vapro.php (manufacturer page: https://www.elitechgroup.com/biomedical-systems/product/vapro-5600/)
  6. Walker JA, Schwartzbard A, Krauss EA, Sherman RA, Eisinger RP. The missing gap: a pitfall in the diagnosis of alcohol intoxication by osmometry. Arch Intern Med. 1986;146(9):1843–1844. doi:10.1001/archinte.146.9.1843. PMID 3753127. https://pubmed.ncbi.nlm.nih.gov/3753127/
  7. Draviam EJ, Custer EM, Schoen I. Vapor pressure and freezing point osmolality measurements applied to a volatile screen. Am J Clin Pathol. 1984;82(6):706–709. doi:10.1093/ajcp/82.6.706. PMID 6507382. https://pubmed.ncbi.nlm.nih.gov/6507382/
  8. Purssell RA, Lynd LD, Koga Y. The use of the osmole gap as a screening test for the presence of exogenous substances. Toxicol Rev. 2004;23(3):189–202. doi:10.2165/00139709-200423030-00005. PMID 15862085. https://pubmed.ncbi.nlm.nih.gov/15862085/
  9. Purssell RA, Pudek M, Brubacher J, Abu-Laban RB. Derivation and validation of a formula to calculate the contribution of ethanol to the osmolal gap. Ann Emerg Med. 2001;38(6):653–659. doi:10.1067/mem.2001.119455. PMID 11719745. https://pubmed.ncbi.nlm.nih.gov/11719745/
  10. Lim M, Dibble A. Osmolality of antisense oligonucleotide parenteral formulations: implications on counterion dissociation and recommended osmometry techniques. Int J Pharm. 2016;515(1–2):788–799. doi:10.1016/j.ijpharm.2016.10.055. PMID 27989829. https://pubmed.ncbi.nlm.nih.gov/27989829/
  11. Hugo Silva M, Hudson SP, Tajber L, et al. Osmolality of excipients for parenteral formulation measured by freezing point depression and vapor pressure: a comparative analysis. Pharm Res. 2023;40(7):1709–1722. doi:10.1007/s11095-022-03262-6. PMID 35460023. https://pubmed.ncbi.nlm.nih.gov/35460023/
  12. Gonotec. OSMOMAT® 050 colloid osmometer: technical data (product datasheet, distributor copy). https://labteamet.com/wp-content/uploads/labteamet/Gonotec/Produktblad_Gonotec_Osmomat_010_070_090.pdf
  13. Morissette MP. Colloid osmotic pressure: its measurement and clinical value. Can Med Assoc J. 1977;116(8):897–900. PMID 851930. https://pmc.ncbi.nlm.nih.gov/articles/PMC1879402/
  14. United States Pharmacopeia. General Chapter 〈785〉 Osmolality and Osmolarity (PF proposal briefing). Rockville, MD: USP; 2024. doi:10.31003/USPNF_M99580_30201_01. https://doi.usp.org/USPNF/USPNF_M99580_30201_01.html
  15. Pang MJ, Habib H, Wang MW, Fang WJ. Non-ideal osmolality behavior in high-concentration protein formulations. Int J Pharm. 2026;690:126553. doi:10.1016/j.ijpharm.2026.126553. PMID 41519180. https://pubmed.ncbi.nlm.nih.gov/41519180/
  16. Clinical Lab Products. Advanced Instruments displays multisample micro-osmometer (OsmoPRO). 2017. https://clpmag.com/diagnostic-technologies/clinical-chemistry/osmometer/advanced-instruments-displays-multisample-micro-osmometer/
  17. Kuang SY, Li X, Yang X, Jones E. A unified framework for van ‘t Hoff’s law: addressing the complexity of osmotic concentration. R Soc Open Sci. 2025;12(9):250622. doi:10.1098/rsos.250622. PMID 40959416. PMC12434622. https://doi.org/10.1098/rsos.250622

Sources

Numbered references are listed in the article above.