This article in 30 seconds
- “99% purity” changes with how you measure. Measured by two methods, the same samples differed by up to 5.2 percentage points.
- The same purity does not mean the same contents. Two samples in the 97% range carried 3 and 7 impurity species respectively.
- The α-form, inactive in the body, never shows up in the number. Chemical synthesis can leave an α-form that shares its molecular formula with β-NMN (the substance the body converts into NAD⁺), so check the manufacturing process and whether content is stated as β-NMN.
Below we take these three in turn, as the three blind spots behind “99% purity.” The values in Fig. 1 and Fig. 2 come from measured data on four samples we obtained.
Change the method and the same sample shifts by up to 5.2 points
A purity figure can only be read together with the method behind it
To start with the conclusion: a “99% purity” with no measurement method stated is not a yardstick for comparison. The number changes with how the sample is measured.
Two methods are mainly used to measure NMN purity: LC-UV (high-performance liquid chromatography with UV detection, which quantifies components by their ultraviolet absorption) and NMR (nuclear magnetic resonance, which probes molecular structure using a magnetic field).
The two see different things. LC-UV is like lighting a dark warehouse with a single flashlight: components that respond to that “light” — ultraviolet — are visible, but salts, solvents and water, which show no absorption in the UV range, are never lit up, however much of them is there. NMR, by contrast, takes stock of the whole warehouse, broadly picking up hydrogen-containing components.
Measuring the same four samples by both methods gave the following.
Sample A reads around 99% either way. Sample B, however, measures 97.2% by LC-UV and 92.0% by NMR — a gap of 5.2 points. Sample C differs by 3.9 points.
The key point
This gap is not measurement error. It means components the flashlight (LC-UV) never saw are visible once you take stock (NMR). The wider the gap between the two numbers, the more of “something” without UV absorption the sample may contain.
So when you set one “99% purity” against another, first confirm which method produced each number. Numbers from different methods were never on the same footing to begin with.
A 0.3-point difference in purity, more than twice as many impurity species
Purity says nothing about what the remaining few percent are
The second blind spot: purity only tells you what percentage is not NMN. What those few percent consist of is an entirely separate piece of information.
For the same four samples, here is the number of impurity species detected by LC-UV.
Sample B: 97.2% purity, 3 impurity species. Sample D: 97.5%, 7 species. The purity difference is only 0.3 points, yet more than twice as many impurity species were detected. Ranked by purity alone, the two look like near-equivalent materials.
The key point
Two materials with similar purity are not equivalents. They are two materials, each carrying its own unexamined few percent. Rather than the purity figure itself, check what was found in that remainder, and how many species.
The α-form, inactive in the body, appears in neither the purity figure nor the impurity list
Where the impurities come from is largely decided by the process
The clue to the third blind spot lies in the manufacturing process, because what the impurities are is decided by how the material is made. Manufacturing methods for NMN fall broadly into three types.
| Item | Full chemical synthesis | Yeast fermentation | Enzymatic biocatalysis |
|---|---|---|---|
| How it is made | Assembled through chemical reactions | Produced by yeast | Converted by enzymatic reaction |
| Stereochemical control | α-form can be present | β-form only | β-form only |
| Expected impurities | Residual reagents and solvents, α-form | Components derived from the culture | Components derived from substrates and enzymes |
| Points to check | Residual solvents; content as the β-form | Removal steps for culture-derived components | Residual unreacted substrate |
This table does not rank the methods. Each process implies different expected impurities, so the points to verify differ accordingly.
Here a difference emerges that shows up neither in the purity figure nor in the impurity list: the α-form, in the “stereochemical control” row of the table.
Right hands and left hands — molecules have handedness too
The shortest route to understanding the α-form is your own two hands. In a mirror, your right and left hands look identical. Try to superimpose them, though, and they never match. A left hand does not fit a right-handed glove.
The same thing happens with molecules. There are pairs of molecules whose atoms are identical in kind, number, and connectivity — differing only in their three-dimensional arrangement. These are called stereoisomers. Mirror-image pairs are enantiomers; pairs differing at only part of their stereochemistry are diastereomers.
And the enzymes in our bodies tell these apart strictly. Like the glove, a molecule whose shape does not fit does not work.
What the β in “β-NMN” refers to
The β written in product names and specifications as “β-NMN” is precisely the symbol for this three-dimensional form. NMN exists as a β-form and an α-form, and the two are anomers (isomers differing in stereochemistry at a single position on the sugar) — differing only at one position on the ribose.
It is the β-form that is produced in the body and converted onward to NAD (a coenzyme involved in cellular energy metabolism). The α-form does not exist in nature; it arises alongside the β-form during chemical synthesis and does not enter the NAD conversion pathway. Peer-reviewed papers likewise treat it as α-NMN lacking physiological activity.
Full chemical synthesis struggles to control this stereochemistry, so the α-form ends up in the product. Processes based on enzymatic reactions or fermentation are stereospecific, yielding the β-form only.
The key point
The α-form has the same molecular formula as the β-form. It therefore never appears as a separate entry in an impurity list organized by molecular formula. Even at “99% as NMN,” how much of that is the β-form is separate information. Two things give it away: the manufacturing process, and content stated as β-NMN.
Three things to check on a spec sheet
The three blind spots above translate directly into three questions for a supplier. When comparing raw materials, ask them in this order.
-
Which measurement method produced the purity figure?
If the number comes from LC-UV alone, components without UV absorption are not reflected in it. A purity figure with no method stated cannot meaningfully be set beside another supplier’s number.
-
What is measured besides purity?
Water content, residual solvents, heavy metals, endotoxin (a fever-inducing substance derived from bacteria) — the number and content of the test items are the real measure of information, because the purity figure says nothing about what the remaining few percent are.
-
What is the manufacturing process, and is content stated as β-NMN?
This is where you learn whether the α-form can be present. Processes using enzymatic reactions or fermentation are stereospecific and yield the β-form only.
For our β-NMN, in addition to purity testing, we publish water content, four heavy-metal tests, and endotoxin results on the product page. For questions about the manufacturing process or measurement conditions, please contact us.
About this data
Figures 1 and 2 are based on our own analysis of four samples we obtained. They show the results for those four specific samples, not trends across all commercially available NMN raw materials. Sample names and countries of manufacture are not disclosed.
The intent is not to rank individual samples, but to illustrate two relationships: purity figures change with the measurement method, and similar purity can hide different numbers of impurity species. Please use them as points of reference when comparing raw materials.
Frequently Asked Questions
Is “purity ≥ 99%” on its own sufficient?
No — that figure alone is not enough to judge. The number changes depending on the measurement method, and purity only tells you what percentage is not NMN; what the remainder consists of is separate information. Check the measurement method together with test items beyond purity, such as water content, heavy metals, and endotoxin.
Between LC-UV and NMR, which figure is correct?
Neither is the correct one — they see different things. LC-UV can only detect components that absorb in the UV range, while NMR broadly picks up hydrogen-containing components. A large gap between the two figures suggests the sample may contain components without UV absorption.
How can I tell whether the α-form is present?
Because the α-form shares its molecular formula with the β-form, it does not appear as a separate entry in impurity lists organized by molecular formula. The practical approach is to check the manufacturing process: enzymatic and fermentation-based processes are stereospecific and yield the β-form only. Also check whether the specification states content as β-NMN rather than as NMN.
This article explains analytical and manufacturing methods for raw materials. Our β-NMN is supplied for research use only and is not intended to demonstrate any specific effect of ingestion.
View β-NMN specifications and test results Is getting NMN from food realistic?