I was told by my chemistry teacher that $\ce{HCN}$ smells like almonds. She then went on to tell a story about how some of her students tried to play a prank on her by pouring almond extract down the drain to make her think that they had inadvertently created $\ce{HCN}$ gas. She said that she knew that it wasn't $\ce{HCN}$ because if she had smelled the almond scent, then she would have already been dead.

I never asked her, but how do people know $\ce{HCN}$ smells like almonds if they would die before they knew what it smells like?

  • 24
    $\begingroup$ That's an exaggeration. Some compounds would probably kill you before you have the chance to feel their smell, but $\ce{HCN}$ would not. $\endgroup$ – Ivan Neretin Mar 1 '16 at 5:46
  • 5
    $\begingroup$ I have a compound in my synthesis right now that smells like almonds. I instinctively am weary of the smell, despite having never (knowingly) smelled HCN. In addition to what everyone else has said, it should be noted that bitter almond oil (apparently) actually contains HCN. I knew amygdalin was a component, but I found out a while back that, supposedly, there is free HCN in the oil (likely a decomposition product of amygdalin). See "side effects" section here. $\endgroup$ – SendersReagent Mar 1 '16 at 7:56
  • 1
    $\begingroup$ @DGS Very interesting! Reminds me in the question by a writer on amygdalin a while ago: chemistry.stackexchange.com/questions/24528/… $\endgroup$ – Klaus-Dieter Warzecha Mar 1 '16 at 8:41
  • 9
    $\begingroup$ As to the comment by your teacher, the smell of the almond extract was probably quite pronounced, so perhaps her comment was along the lines of, "that much HCN would have produced symptoms already, so it isn't HCN". $\endgroup$ – DevSolar Mar 1 '16 at 10:42

The odour threshold for HCN is in fact quite a bit lower than the lethal toxicity threshold. Data for hydrogen cyanide can be found in many places, but here and here are a couple of good references. That subset of the human population that can detect bitter almonds do so at a threshold of 0.58 to 5ppm. The lethal exposure dose is upwards of 135ppm. That's a whole 100ppm range in which to detect and report the fragrant properties.

  • 15
    $\begingroup$ You did beat me by the length of a shower and a shave :D Further data on inhalation effects can be found in the SCOEL Recommendation 115. One should however keep in mind that a number of people can't smell HCN at all. $\endgroup$ – Klaus-Dieter Warzecha Mar 1 '16 at 6:50
  • 9
    $\begingroup$ I wonder who was the first to volunteer for this experiment. "100ppm and you won't die, I swear, just sniff this jar and tell me what you smell!" $\endgroup$ – hownowbrowncow Mar 1 '16 at 15:39
  • 6
    $\begingroup$ One way to look it at this is that our olfactory system wouldn't have been much of an evolutionary advantage if it didn't warn us about poisons before they killed us. $\endgroup$ – Crashworks Mar 2 '16 at 0:47
  • 2
    $\begingroup$ @Crashworks True! But also why we can make use of sentinel species like the canary in the coalmine :-) $\endgroup$ – long Mar 2 '16 at 0:59
  • 6
    $\begingroup$ Missing a major point: inhaling 100 ppm for a short time has absolutely no ill effects, as you need to accumulate a certain amount (about 10 mg) of cyanide in your blood for it to kill you. One breath would have to be highly concentrated to kill you - on the order of 10,000 ppm. See my answer for details. $\endgroup$ – Floris Mar 2 '16 at 1:41

Recognize that a whiff of most toxins, even in high concentration, will probably not kill you. You need a sufficient concentration in your blood - which means you have to actually get a certain number of HCN molecules to penetrate across the mucosa of the lung and into the blood stream. Typical breathing volume is about 500 mL (tidal volume), about 1/40th of a mole, so a single normal breath of 100 ppm HCN would contain 2.5 µmol of HCN.

According to the CDC, humans can tolerate inhalation of 50 ppm HCN for half an hour "without immediate or delayed effects", while 100 ppm for more than half an hour may be fatal.

The Mayo clinic reports that blood cyanide levels over 2 µg / ml are toxic; if you have 5 liters of blood, that corresponds to 10 mg. The molar mass of HCN is 27, meaning that 10 mg is 370 µmol. At 2.5 µmol / breath, you should be able to take 148 breaths; at 6 breaths per minute, that would be about 25 minutes.

All these numbers are quite consistent, and it tells you that you can tolerate a whiff of cyanide without ill effects. But it's a good idea, once you smell it, to open the windows, turn on the extractor fans, and get yourself to some fresh air.


Hydrogen cyanide $(\ce{HCN})$ is variously described as smelling of bitter almonds, marzipan, ratafia, or peach kernels. While some people can smell $\ce{HCN}$ at very low concentrations, many people cannot perceive the odour at all. The odour threshold is about $1{-}6\ \mathrm{mg/m^3}$ for people who are actually sensitive to the odour of $\ce{HCN}$.

Inhalation of $\ce{HCN}$ at low concentrations above the odour threshold is not necessarily lethal. Toxicity of inhaled cyanides is strongly dependent upon concentration and exposure time. By way of comparison, guideline figures taken from Marrs, T. C.; Maynard, R. L.; Sidell, F. R. Chemical warfare agents: toxicology and treatment; John Wiley & Sons, 1996; p 204 are shown in the following table:

$$\textbf{Inhalation toxicity to humans}\\ \begin{array}{lll} \hline \text{Time} & \mathrm{LC_{50}} & \mathrm{LCt_{50}} \\ \text{in}\ \mathrm{min} & \text{in}\ \mathrm{mg\ m^{-3}} & \text{in}\ \mathrm{mg\ m^{-3}\ min} \\ \hline \hphantom{0}0.25 & 2\,400 & \hphantom{0\,}660 \\ \hphantom{0}1 & 1\,000 & 1\,000 \\ 10 & \hphantom{0\,}200 & 2\,000 \\ 15 & \hphantom{0\,}133 & 4\,000 \\ \hline \end{array}$$

Hence, for example, it is possible to smell $\ce{HCN}$ for a period of $1\ \mathrm{min}$ at a concentration of $100\ \mathrm{mg/m^3}$, which is well above the odour threshold $(1{-}6\ \mathrm{mg/m^3})$ but well below the lethal concentration for this period $(1\,000\ \mathrm{mg/m^3})$. However, after an extended period, inhalation of $\ce{HCN}$ at this concentration will probably become lethal.

Even at acute exposure to high concentrations of $\ce{HCN}$, it is possible to smell $\ce{HCN}$ before the toxic effects occur. This may be illustrated by a description given in Vedder, E. B. The Medical Aspects of Chemical Warfare; Williams and Wilkins, 1925; p 187:

In an atmosphere containing a lethal concentration an odour of bitter almonds is noticed. This is followed by a sensation of constriction of the throat, giddiness, confusion and indistinct sight. The head feels as though the temples were gripped in a vice, and there may be pain in the back of the neck, pain in the chest, with palpitation and laboured respiration. Unconsciousness occurs and the man drops. From this moment if the subject remains in the atmosphere of hydrocyanic acid for more than two or three minutes death almost always ensues, after a brief period of convulsions followed by failure of respiration.

However, the figures given above are extremely uncertain. Nevertheless, note that $\ce{HCN}$ does not obey Haber’s rule $(c \cdot t = k)$. One important reason for the dependency of the toxicity upon concentration is the existence of various pathways for detoxication. Detoxication explains the ability to withstand very low concentrations of cyanide indefinetely. However, it is unlikely that detoxication plays any significant role in acute cyanide poisoning.

A significant contribution to the uncertainty of acute inhalation toxicity is caused by the variable breathing rate. Standard reference values taken from ICRP 66 (1994) are shown in the following table:

$$\textbf{Breathing rates for adult males}\\ \begin{array}{lll} \hline \text{Acitivity} & \text{Breathing rate} \\ & \text{in}\ \mathrm{m^3\ h^{-1}} \\ \hline \text{Resting (sleeping)} & 0.45 \\ \text{Sitting awake} & 0.54 \\ \text{Light exercise} & 1.5 \\ \text{Heavy exercise} & 3.0 \\ \hline \end{array}$$

However, such values may not be applicable to acute cyanide poisoning because of the respiratory stimulation caused by $\ce{HCN}$.


Gatterman reports (Org. Synth. 1927, 7, 50, as a footnote) that people who smoke regularly have enhanced sensitivity to the smell of cyanide gas, and he recommend smoking while preparing it!

Organic Synthesis Collective Volume 1 1941 314-315

Just opening the NaCN container, most regular (and former regular) smokers can smell the trace amount of HCN formed from the water vapor in the air.

I suspect your teacher was exaggerating slightly.

  • $\begingroup$ Even if cyanide gas isn't a reactive fuel for a cigarette butt, surely other gases in the lab are -- that doesn't sound like sound advice to me :P $\endgroup$ – cat Mar 2 '16 at 0:17
  • 8
    $\begingroup$ It isn't sound advice, the recommendation is from 1927. Old-fashioned, outdated and dangerous. I wasn't smoking when I was making hydrogen cyanide. $\endgroup$ – Lighthart Mar 2 '16 at 15:28
  • $\begingroup$ It may not be sound advice, but it is true that smokers smell cyanide at much lower levels. It is probably one of the few positive benefits of smoking. $\endgroup$ – matt_black Mar 3 '16 at 17:12

how do people know HCN smells like almonds

Bit late to the party, but I'm missing a crucial part in the existing answers. So here are my 2ct: HCN does not smell like [bitter] almonds.

Personally, I find the description "HCN smells like almonds" very confusing: what we typically associate with the smell of almonds is benzaldehyde rather than HCN.

I believe that confusion comes from crime stories probably written by people who knew that bitter almonds "contain" cyanide (and that [some] people can smell cyanide) and also knowing the smell of bitter almonds (after all, they are/were used for their arome) - but not knowing/realizing that the major arome component is not the cyanide.

Bitter almonds contain amygdalin, a cyanogenic glucosid. Amygdalin can be hydrolysed into gentiobiose (disaccharid of 2 x glucose) and mandelonitrile, the nitrile of mandelic acid (Mandel = German for almond) which in turn is hydrolyzed into HCN and benzaldehyde. This leaves us with 2 volatile compounds, benzaldehyde and HCN.

  • Benzaldehyde has a very strong smell (odor threshold 0,2 mg/m3 ≈ 0.04 ppm).

  • In addition, also the HCN can be smelled (not by everyone, btw. but by many people, and with varying detection limits - this has been studied including the genetics in detail in the 1950s/60s, odor threshold 0.2-5 ppm)

  • Bitter almods are used as spice for baking, but keep in mind that HCN is not very stable (e.g. you cannot keep small amounts of KCN or NaCN very long at ambient air: they will be oxidized into cyanate). The HCN basically doesn't make it into the baked cookies.
    What we associate with "almond" is the benzaldehyde.

  • Also the artificial "almond arome" you can buy (at least here in Germany) is benzaldehyde without HCN. (An oil exctracted from bitter almonds or other prunus kernels may contain or form HCN, though)

Personal experimental anecdata:

  • I believe I'm a reasonably good smeller of HCN (I did solve that part of a fellow student's anion analysis in our first semester that way by sniffing, and comparing to the smell over KCN salt)

  • Still, mashing bitter almonds in water to me leads to an overwhelming benzaldehyde smell that masks the HCN smell.
    I do smell the HCN, though, if the bitter almonds are mashed in a (non-smelling) acidic solution (e.g. dilute sulfuric acid): that way, HCN smell comes first before benzaldehyde takes over.

So the chemistry teacher could (should!?) have known that the joke smell was benzaldehyde.
(And that the risk of expensive natural bitter almond oil being used instead of cheap benzaldehyde was minute - not to say non-existent given that the students were doing well ;-))


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.