HPLC testing laboratory · Davis, CaliforniaPublished pricesBook onlineinfo@openchemworks.com
Insights

Ammonia: The Number Nobody Ordered but Everybody Gets

Published November 5, 2026Updated November 19, 2026OpenChemWorks Laboratory

Updated November 19, 2026

Summary: key findings

  • Glutamine in stored media decomposes on a clock you can compute. Tritsch & Moore (1962), as reported in a 2024 bioRxiv review of glutamine decomposition in cell culture media, found first-order decomposition with 10% conversion in 1 day at 37 °C and in 9 days at 4 °C. From those two anchor points: the rate constant is k = 0.105/day at 37 °C (half-life 6.6 days) and k = 0.0117/day at 4 °C (half-life 59.2 days). Every number in this article is computed from those anchors, with the arithmetic shown.
  • A 4 mM glutamine medium held at 37 °C crosses 2 mM ammonia in about 6.6 days from chemistry alone, with no cells present. Ammonia(t) = 4 × (1 − e^(−0.105t)); at t = 6.58 days, ammonia = 2.00 mM. At 14 days it reaches 3.09 mM, and at 30 days 3.83 mM.
  • Refrigeration slows the clock a lot, but it does not stop it. At 4 °C, a 4 mM medium reaches 1.19 mM ammonia at 30 days, 2.02 mM at 60 days, and 2.61 mM at 90 days. The 4 °C anchor is one published point. Khan and co-workers report glutamine loss under 0.2% per day at 4 °C, which is slower than the 1.2% per day this anchor implies, so read the 4 °C numbers as an upper bound and expect real refrigerated media to sit below them. Which one your medium follows depends on its composition and pH.
  • Storage alone cannot push a standard medium past the classic CHO inhibition threshold, and that is precisely the trap. Yang & Butler (2000) found CHO growth inhibited above 5 mM ammonium chloride, IC50 33 mM. A 2–4 mM glutamine medium can never exceed 2–4 mM ammonia from decomposition alone, because ammonia cannot exceed the glutamine it came from. But sensitive lines are inhibited far lower (hybridoma IC50 values of 3–7 mM are reported in the literature Yang & Butler review, and Hassell et al. reported IC50 values as low as 0.8 mM across nine mammalian lines), and in culture, cellular glutamine metabolism stacks on top of the storage load. Media that arrives pre-loaded with 1–3 mM ammonia starts the culture that much closer to the line.
  • Published rates disagree by about 2×, and the disagreement is information. A second anchor, Souba et al. (1985), reported about 5% glutamine loss per day at 37 °C in parenteral solution (k = 0.051/day, half-life 13.5 days), roughly half the Tritsch & Moore rate. Ouyang et al.'s work on glutamine decomposition in media preparations shows why: the rate constant rises with pH (studied over pH 6.8–7.8 at 37 °C) and depends on the medium. Both curves are tabulated below. Use the fast curve for planning; use the spread as your uncertainty.

So what: Your medium's ammonia content on the day you use it is a function of its glutamine concentration, its storage temperature, and its age, and all three are knowable before you open the bottle. If your process is ammonia-sensitive, the acceptance test is cheap and the failure it prevents is not.

Why it matters: Ammonia is reported on every amino acid panel we run, including spent media, whether the customer asked for it or not. Read against a sample's history, it is a fingerprint: fresh sample, aged sample, or a culture that was already fighting its own medium before the cells added anything.

Introduction

Glutamine dissolved in stored media decomposes on its own, with no cells present, and every mole that decomposes releases a mole of ammonia. This article computes how much ammonia that chemistry alone accumulates, as a function of the three things a buyer actually knows before opening the bottle: the starting glutamine concentration, the storage temperature, and the age of the medium.

It matters because ammonia is reported on every amino acid panel, whether the customer asked for it or not, and because sensitive cell lines are inhibited at ammonia levels far below the classic thresholds for robust production lines. A medium that arrives already loaded with ammonia from storage starts the culture that much closer to the line, before a single cell has metabolised anything.

The curves are computed from published anchors for first-order glutamine decomposition, reported for refrigerated and incubator temperatures, with a second, slower published rate carried through as a sensitivity check; the disagreement between the two rates is itself part of the finding. The inhibition thresholds stacked against the curves are drawn from the published cell-culture literature reviewed in the thresholds section.

Read the warm curve as the planning case and the spread between the fast and slow curves as the uncertainty. Then compare the accumulated ammonia at your medium’s actual age against the threshold for your own cell line, not against a generic one.

Ammonia accumulated at 37 °C

Cell-free decomposition of glutamine, fast curve. Ammonia cannot exceed the glutamine it came from.

Starting glutamine 4 mMStarting glutamine 2 mM
0123407.51522.530daysammonia accumulated, mM2 mM ammonia3.831.92
Figure 1: At 37 °C the medium starting with 4 mM glutamine crosses the 2 mM ammonia line; the medium starting with 2 mM approaches the same line but cannot cross it, because ammonia cannot exceed the glutamine it came from.

The computed curves

Model: first-order decomposition, Gln = [Gln]0 × e^(−kt); ammonia(t) = [Gln]0 × (1 − e^(−kt)), assuming 1 mol ammonia per mol glutamine decomposed (decomposition to pyrrolidonecarboxylic acid plus ammonia, per the literature cited). k derived from Tritsch & Moore (1962): 10% conversion means e^(−kt) = 0.90, so k = −ln(0.90)/t = 0.10536/day at 37 °C (t = 1 day) and 0.10536/9 = 0.01171/day at 4 °C.

At 37 °C (k = 0.105/day)

Days NH3 from 2 mM Gln (mM) NH3 from 4 mM Gln (mM) NH3 from 4 mM, slow curve k = 0.051/day (mM)
1 0.20 0.40 0.20
3 0.54 1.08 0.57
7 1.04 2.09 1.21
14 1.54 3.09 2.05
30 1.92 3.83 3.14

At 4 °C (k = 0.0117/day)

Days NH3 from 2 mM Gln (mM) NH3 from 4 mM Gln (mM)
7 0.16 0.32
14 0.30 0.61
30 0.59 1.19
60 1.01 2.02
90 1.30 2.61
180 1.76 3.51

Worked example, time to 2 mM ammonia for a 4 mM medium at 37 °C: solve 2 = 4(1 − e^(−0.105t)) → e^(−0.105t) = 0.5 → t = ln(2)/0.105 = 6.6 days. The same threshold at 4 °C: t = ln(2)/0.0117 = 59.2 days. (This is also the glutamine half-life at each temperature, by the symmetry of the equation.)

Arii et al. (1999), studying glutamine degradation kinetics in aqueous solution, found pseudo-first-order degradation, maximum stability at pH 5.0–7.5, and an apparent activation energy of 9.87 × 10^4 J/mol at pH 6.41. That activation energy is why the 4 °C and 37 °C rates differ by an order of magnitude, and why a summer shipping lane at ambient temperature ages media far faster than the refrigerator arithmetic suggests. This version does not tabulate ambient-temperature curves, because no anchor rate at 20–25 °C was verified from a fetched primary source in this pass; the two verified anchors are 4 °C and 37 °C.

Ammonia accumulated at 4 °C

Refrigeration slows the same chemistry by about ninefold; it does not stop it.

Starting glutamine 4 mMStarting glutamine 2 mM
0123404590135180daysammonia accumulated, mM2 mM ammonia3.511.76
Figure 2: At 4 °C the medium starting with 4 mM glutamine still crosses the 2 mM ammonia line, on a timescale of months rather than days.

Thresholds, stacked against the curves

Threshold Value Source
CHO growth inhibition begins above 5 mM NH4Cl Yang & Butler (2000), Biotechnol Bioeng 68:370–380
CHO IC50 33 mM NH4Cl Yang & Butler (2000)
CHO 50% growth reduction above 8 mM ammonia Reported in CHO metabolism reviews citing Yang & Butler-era data
Hybridoma IC50 range 3–7 mM Doyle & Butler (1990) and Dodge et al., reviewed in Yang & Butler (2000)
Mammalian lines, IC50 range (9 lines) 0.8 to >5 mM Hassell et al. (1991), reviewed in Yang & Butler (2000)

The honest reading: for a robust CHO process with a 5 mM concern threshold, storage decomposition alone (max 4 mM from a 4 mM medium, reached only asymptotically) is a contributor, not the whole story. For a sensitive line with an IC50 near 1–3 mM, a month-old warm medium or a quarter-old refrigerated medium can arrive already inside the effect range, before a single cell has metabolised anything.

Perspectives

Process. Shelf-life and freeze-thaw decisions follow from the curves: aliquot and freeze glutamine separately where the process allows, date media by glutamine-addition date rather than manufacture date, and treat ambient shipping time as 37 °C-class aging for planning purposes until ambient anchor data is added.

Analytical. An ammonia value on a spent-media report is a sample-history fingerprint. High ammonia with high residual glutamine suggests cellular metabolism; high ammonia with depleted glutamine in a cell-free or early sample suggests age and storage. The pair (Gln, NH3) is more diagnostic than either alone, which is why our panel reports ammonia on every run.

Formulation. Stabilised glutamine dipeptides (alanyl-glutamine, ) exist precisely because this chemistry is unavoidable for free glutamine; they release glutamine enzymatically and do not accumulate ammonia spontaneously at these rates. Price them against the decomposition math above: the cost of the dipeptide versus the cost of the ammonia load your storage pattern guarantees.

QC. Incoming-media acceptance can carry numbers: for a 4 mM glutamine medium, an ammonia acceptance limit of 0.5 mM at receipt corresponds, on the fast curve, to about 1.3 days at 37 °C-equivalent aging or at most about 11 days at 4 °C. Set the limit from your line's sensitivity (table above), not from habit.

Worked example for the QC limit: 0.5 = 4(1 − e^(−kt)) → e^(−kt) = 0.875 → t = 0.1335/k. At k = 0.105: t = 1.3 days. At k = 0.0117: t = 11.4 days.

What to do with this

Measure it. Glutamine and ammonia are both on the AA-1 panel, so one run compares your storage arithmetic with what is really in the bottle or the flask.

What we can and can't measure for this question

AA-1 ($150 per sample) reports glutamine and ammonia in µM and mg/L, from 200 µL of medium (400 µL preferred). For a bioreactor time course, RUN-1 is $1,990 for 24 time points and shows when ammonia builds and glutamine runs down. We do not model your medium for you: the curves above are cell-free chemistry from published rates, and your own data should win over them. Research use only.

Order a time course Order single samples See an example report

Overnight turnaround available, with expedited service on request. Standard results in 3 to 5 business days from receipt.

Methods & Sources

  • Kinetic anchors: Tritsch & Moore (1962), first-order glutamine decomposition in cell-free media, 10% conversion in 9 days at 4 °C and 1 day at 37 °C, as reported in "Decomposition of L-glutamine and accumulation of ammonium in cell culture media inhibit infectivity of influenza viruses", bioRxiv 2024, doi:10.1101/2024.06.21.600004, fetched in full text 2026-11-05. The primary 1962 paper was not fetched directly; the anchor is cited through this review, and that is stated here rather than hidden.
  • Second anchor: Souba et al. (1985), about 5% per day glutamine loss at 37 °C just below neutral pH, as reported in the abstract of Arii et al. (1999), "Degradation kinetics of L-glutamine in aqueous solution", European Journal of Pharmaceutical Sciences (ScienceDirect abstract, fetched 2026-11-05). Arii et al. also supply the pH 5.0–7.5 stability maximum and the activation energy 9.87 × 10^4 J/mol at pH 6.41.
  • pH and medium dependence of the rate constant at 37 °C: Ouyang et al., "Chemical decomposition of glutamine in cell culture media" (abstract and rate table via Academia.edu, searched 2026-11-05): first-order constants increase with pH over 6.8–7.8; serum concentration had little effect; medium type did.
  • Thresholds: Yang & Butler (2000), Biotechnol Bioeng 68:370–380, via the published abstract and full text copies fetched 2026-11-05; CHO metabolism review (Biotechnology Journal, doi:10.1002/biot.201700499) for the 5.1 mM CHO inhibition and >8 mM 50% reduction figures.
  • Stoichiometry assumption: 1 mol ammonia per mol glutamine decomposed (to pyrrolidonecarboxylic acid + ammonia). Cellular ammonia production is additional and is not modelled here.
  • LIMITS: the model is cell-free chemical decomposition only, at two verified temperatures, from anchors reported through secondary sources as noted. It does not predict any specific commercial medium's behaviour, because medium composition and pH move the rate constant (Ouyang et al.). Measure your medium; do not litigate the model.
Order builder

Have samples in mind?

Pick the tests, the sample type and any preparation. The total updates as you go, and you can book in the same place. Every report includes the chromatogram and QC behind the numbers.

For the short version, see how much does amino acid testing cost: the per-sample numbers, and why quotes vary.

  • Free amino acids$150
  • Total amino acids$230
  • Creatine$120