Threshold Data

mixmastermattyp

Well-known member
Nov 11, 2020
For those who don't know: the odor threshold is the minimum concentration of which an aroma chemical can be detected by our noses. It's an important quantity because perceived intensity is inversely proportional to the odor threshold:

I = (C/T)^s

where I is the perceived intensity of an aroma chemical, C is the concentration of the aroma chemical in the vapor phase, T is the odor threshold of the aroma chemical, and s is the Steven's Power Law exponent which is associated with the aroma chemical's slope of psychophysical function.

It's not 100% essential to understand the formula. The main point I'm concerned with at the moment is that the higher the odor threshold, the more of an aroma chemical you have to use to increase the perceived intensity. For example: a user posted a while back a link to a nice table with some data on aroma chemicals contained in rose (http://www.leffingwell.com/rose.htm). Looking at the table, it says that the odor threshold of Phenethyl Alcohol is 750 ppb. The threshold of linalool is 6 ppb. That means, if you want to make an accord which smells roughly equal of phenethyl alcohol and linalool, you would need to use 125 times as much phenethyl alcohol as that of linalool.

That's why we do the Jean Claude type experiments testing each material in successive ratios of concentration 1:10, 2:10, etc. etc. But knowing how widely these thresholds vary, its easy to see how impractical these types of experiments can be. Of course, having a general idea of how strong materials are can expediate experimentation, but its still tedious to say the least.

So my question is: Does anybody have a source of data for the odor thresholds of aroma chemicals like the one i linked to? I'm familiar with the data collected by symrise (https://www.symrise.com/scent-and-care/aroma-molecules/ingredient-finder/), but its fairly limited. Is there any more data available anywhere?

Thanks y'all.
 

apolo085

Well-known member
Feb 18, 2019
I was thinking about this many months ago.
There is two types of odour threshold values, the odor detection threshold which is the minimum mandatory value to detect the presence of the compound in the headspace, and the odor recognition threshold which is the value which allows a complete recognition of the compound in the headspace.
I was always confused on the rare records I found about odor threshold values, what type of threshold are we talking about? I think this is an important point to focus on.
I think I may have seen a threshold provider, but I am recalling it being not free.

Also this must be wrong:
Looking at the table, it says that the odor threshold of Phenethyl Alcohol is 750 ppb. The threshold of linalool is 6 ppb. That means, if you want to make an accord which smells roughly equal of phenethyl alcohol and linalool, you would need to use 125 times as much phenethyl alcohol as that of linalool.
An azeotropic accord between Linalool/PEA must be around 1:1.67
There is calculations to do to meet that model somewhere (using the odor recognition threshold).
 
Last edited:

mnitabach

Well-known member
Nov 13, 2020
My sense as someone with a decent amount of experience as a sensory physiologist is that as a practical matter for purposes of perfumery this psychophysical function is waaaay too coarse an approximation & there is little or no value to starting from relative odor thresholds (even assuming they have been measured in a consistent way across odorants). Your example, in fact, highlights this point, as PEA & linalool pretty much never appear in any accords or formulas at anywhere near a 125:1 ratio (and even ignoring the fact that you would have to normalize such a ratio derived from relative odor threshold in terms of air concentration to relative vapor pressure). The best way to get a starting point for accord experiments is to look at as many existing formulas as possible that contain simultaneously aromachemicals of interest to get a sense for commonly used ratios among them.
 

Casper_grassy

Well-known member
May 5, 2020
I do not have what you’re looking for, but I can say 1. This seems like a significant waste of time and 2. If you make 125:1 ratio of linalool and pea, you will not smell any pea, therefore off the top of my head I can surmise this is pointless. Sorry to tell you.

All of these stats or whatever you may call them will lend no success. Trust that thing you got in the middle of your face, it works better, I promise.
 

mnitabach

Well-known member
Nov 13, 2020
I do not have what you’re looking for, but I can say 1. This seems like a significant waste of time and 2. If you make 125:1 ratio of linalool and pea, you will not smell any pea, therefore off the top of my head I can surmise this is pointless. Sorry to tell you.

All of these stats or whatever you may call them will lend no success. Trust that thing you got in the middle of your face, it works better, I promise.

I think they were saying the opposite PEA:linalool 125:1, but the point is the same. Even if you normalized for the nominal relative vapor pressures listed on Good Scents & did ~30:1 PEA:linalool, you'd be nowhere near a useful or interesting accord. In fact, there is nearly always a lot more linalool than PEA in any accord or complete fragrance that contains both. Anyways, there are a whole bunch of reasons, including more that we haven't even raised, why this is a cockamamie idea. Every so often, someone posts on here w some supposed magic theory of how to predict useful relative doses of aromamaterials. It's all bullshit.
 

apolo085

Well-known member
Feb 18, 2019
.. Every so often, someone posts on here w some supposed magic theory of how to predict useful relative doses of aromamaterials. It's all bullshit.
No need to be rude on people trying to understand other methods (if exist) to make perfumes, the guy was asking simple questions to understand and was by no mean affirming having discovered a new way to predict what so ever.
You, I or maybe anybody in this forum does not know everything about perfumery, otherwise we will not be in this forum.
Keeping a certain level of doubt is key to perpetual learning in any discipline including this one.

Anyways, there are a whole bunch of reasons, including more that we haven't even raised, why this is a cockamamie idea.
May be exposing these reasons will be more knowledgeable and contributory than just calling all of that: "bullshit"

Just saying...
 

parker25mv

Well-known member
Oct 12, 2016
I = (C/T)^s

where I is the perceived intensity of an aroma chemical, C is the concentration of the aroma chemical in the vapor phase, T is the odor threshold of the aroma chemical, and s is the Steven's Power Law exponent which is associated with the aroma chemical's slope of psychophysical function.
That is probably overall generally mostly true, but not all odor substances behave this way.

Some substances, like damascones, can be detected in very minute quantities, but also do not become overbearing when present at substantially higher levels.

the odor threshold of Phenethyl Alcohol is 750 ppb. The threshold of linalool is 6 ppb. That means, if you want to make an accord which smells roughly equal of phenethyl alcohol and linalool, you would need to use 125 times as much phenethyl alcohol as that of linalool.
Even that may not be entirely true and may be a little more complicated. Some substances have certain olfactory aspects to them that are much more potent than other aspects which they possess. So if you use something in very small amounts, you might only be able to smell the very sharp potent aspect of it, when the reality is most of it does not really come into being at those ultra low levels.
 

mnitabach

Well-known member
Nov 13, 2020
That is probably overall generally mostly true, but not all odor substances behave this way.

Some substances, like damascones, can be detected in very minute quantities, but also do not become overbearing when present at substantially higher levels.


Even that may not be entirely true and may be a little more complicated. Some substances have certain olfactory aspects to them that are much more potent than other aspects which they possess. So if you use something in very small amounts, you might only be able to smell the very sharp potent aspect of it, when the reality is most of it does not really come into being at those ultra low levels.

Additional well-explained reasons why this entire idea is a waste of time.
 

parker25mv

Well-known member
Oct 12, 2016
Additional well-explained reasons why this entire idea is a waste of time.
That is definitely not what I said. I think generalities and basic rules describing phenomena can still be very useful to understand what is happening. It can help us fill in the blanks when we do not always have specific information about empirical observations. And the two can even be combined together sometimes, where we can take actual specific observed empirical data and then extrapolate from them.

In terms of mathematics, almost everything is going to look like a decreasing logarithmic curve. They will just have different slopes, different rates of decrease, depending on their substantivities.

Natural materials and accords will of course be more complicated, because there are multiple molecules in them, each with different rates of decrease. The smell will shift over time, occasionally perhaps in a more complex way that is not all entirely linear.
 

Latest News

Top