A full guide to vape aerosols. Post 5: metals

Summary

This is the fifth Substack post of a series of posts describing vaping aerosols, their properties, their optimal regime of operation and comparisons with tobacco smoke and other aerosols. Understanding how vape aerosols form, operate and can be tested provides the knowledge to understand their pleasurable usage, their toxicity profile and relative safety with respect to tobacco smoke and other aerosols and pollutants. Without being “experts” this knowledge reassures our confidence on the role of vapes in harm reduction and serves us to counter ignorant and malicious disinformation.

This knowledge (which i try to present accessibly) reassures our confidence on the role of vapes in harm reduction and serves us to counter ignorant and malicious disinformation.

Previous Substack posts: Post 4, Post 3, Post 2, Post 1.

Post 5: Metals in vape aerosols

I described in previous posts (Post 2 and Post 3) the conditions of aerosol generation in an Optimal and Overheating Regimes, while Post 4 provides a set of guidelines to generate and test vape aerosols in a laboratory. Aerosol generation in the laboratory requires a standardized regimented puffing regime in which user inhalation is simulated by vaping machines.

While regimented puffing does not (can not) reproduce real life puffing, it is necessary to program vaping machines with appropriate puffing parameters that approximate consumer usage as best as possible. It is specially important to avoid vaping machines operating under overheating conditions that generate aerosols repellent to users.

There is a known literature of emission studies that have reported different metallic elements (nickel, lead, copper, manganese) in vape aerosols. Since some metals are carcinogenic, their the presence in vape aerosols is concerning. However, it is important to understand the physical process that explain how can metallic elements be found in the aerosol and e-liquids.

Many emission studies detecting metals have voiced alarm and received wide media covering, but it is necessary and important to verify their experimental design and if their conducted a correct comparison of exposure doses to toxicological safety markers. Sebastien Soulet and I published an extensive review of studies focusing on metal element contant in vape aerosols. All studies reporting high toxicity from metals in vape aerosols exhibit serious methodological flaws and are thus unreliable.

Solids, metals, liquids gases in general

The difference between these forms of matter is in the type of atoms (chemistry), how they bind to each other and the different structures in which they are arranged.

Solids

• Ceramics: molecules (typically metal & non-metal atoms) form tight multiple chrystalline strucures organized in repeatable patterns. Molecules are strongly bound by an exchange and/or sharing of electrons (ionic and covalent bounds). These strong bounds explains their tough hardness, but also their brittle nature.

• Metals. Atoms and molecules in alloys are also tightly bound (ionic bound), but not as strong as in ceramics. So, electrons move freely along the full structure made of positively changed nuclei ions (cations). This explains their high capacity to conduct electricity and heat.

• Polymers form crystaline structures of organic macro-molecules bound by weak forces, which explains their elasticity. and ductibility.

Liquids and glasses

• Molecules move randomly in amorphous distributions, their bounds are intermediate: much stronger than in gases and much weaken than in solids. Glasses are solids whose molecules form amorphous structures similar to liquids, but their atoms are strongly bound.

Gases

• Molecules are distant, randomly distributed, very weakly interacting and very mobile.

Metals and gases in “bulk” (no contact)

Before discussing metals in vape aerosols, let us first compare atoms/molecules when there is no contact: in “pure” bulk metal and “pure” bulk gas.

Figure 1 Bulk metal and bulk gas.

Gas molecules in the pure gas bulk are practically free of interactions until they collide. Forces are small and random. As a contrast, metals are characterized by very tight crystalline atomic structures (see Figure 1), with positively charged atomic nuclei (red circles) close to each other and bound by strong inter-atomic microscopic forces. A metal nucleus in the bulk (green circle) is evenly surrounded in all directions by trillions of other nuclei, hence the atomic forces from other nuclei (green arrows) is cancelled by atomic forces exerted on it from all directions. As a consequence, each nuclei remains firmly in the crystal undergoing small vibrations. The vibration of the nuclei and this shared electron ocean explains why metals are good conductors of electricity and heat..

Metals in vape aerosols

It is not surprising that various metal elements have been detected in vape aerosols, since the heated e-liquid and the vapor-air mix is in contact with a metallic coil made of various alloys (in early ciga-likes, they were also in contact with wire solders). However, metal elements can also be detected in food cooked in heated metallic pans and ovens. In cigarette smoke metals are deposited from the atmosphere on the tobacco leaf and transported during its combustion.

So how do metals occur in vape aerosols? While the process has not been fully researched, surface physics most likely provides the answer, where the term “surface” describes an interface of interaction.

The metallic coil is in contact with the heated e-liquid and with its vapor around its neighborhood. This defines a metal-liquid and a metal-gas interfaces where metal molecules of the coil interact with molecules of the e-liquid and/or its vapor.

The popular notion of a “surface” is that of a very “thin sheet” between thermal states (gas, liquid, solid). However, at a molecular lever there is no sharp boundary, but a microscopically wide (macroscopically thin) transition interface in which very interesting physical process take place.

The near equilibrium of atomic forces in bulk metal (Figure 1) changes for nuclei close to an interface, the macroscopically thin zone of the transition to a liquid or gas. Nuclei close to the interface are attracted by much less nuclei in the direction of the interface than in the much more numerous nuclei in the direction of the bulk. As a consequence, the energy from the vibration in nuclei at the interface might be sufficient to break the bond and escape to the liquid or gas (red curved arrows in Figure 2).

However, the escape through the interface towards a gas is far more likely than towards a liquid, since vibrating metal nuclei will face little resistance from widely spread out gas molecules whose atomic forces are negligible, while inter-atomic forces are stronger among liquid molecules (though much less than forces in the metallic crystal).

How do liquid and gas molecules behave at the interface? Molecules in a liquid that also has an interface with the gas experience a phenomenon called surface desorption (whose macroscopic manifestation is called “wettability”). As they approach the liquid-metal and liquid gas interfaces they tend to form droplets that “stick” as layers into the external part of the metal-gas interface (green arrows in Figure 2). The droplets form a desorption layer on the metal surface sustained mostly by their surface tension (forces generated in the droplet surface that compensate external gas pressure). Since metal nuclei are much heavier than liquid or gas molecules, some might desorb and some might escape into the gas.

Figure 2: the metal bulk and its interfaces

The surface phenomena described before are very sensitive to temperature and pressure. As temperature in the interface increases more metal nuclei acquire sufficient vibrational energy to overcome binding forces and will escape into the gas (even into the liquid). Other factors are the geometry and size of the interface and specially the rugosity of the surface or unevenness due to small scale variations of the interface region. Surface phenomena are much more complicated in porous materials than in metals.

Metal content in vaping aerosol is very likely explained by metallic ions that escape from the metal-gas interface during e-liquid vaporization (much more likely than through the metal to e-liquid interface). As the metallic nuclei (ions) escape they react with oxygen to form metallic oxides, which then should cluster (nucleate) forming very small (nanometer size) solid particles of at most few tens of nanometers (1 thousand million parts of a meter) transported by the aerosol. Evidently, under an Overheating Regime with coil temperatures rapidly increasing above e-liquid boiling temperatures, a much larger amount of metallic ions will escape the interface and form more of these solid particles (see Post 3 for a full discussion on overheating).

Metallic nano-particles

Is there evidence of metallic nano-particles in vape aerosols? How many such particles form? While there is a large literature quantifying total metal element content in the analysis of the aerosol (see further ahead), quantifying actual numbers of these particles has received little attention in the literature. Metals in vape aerosols as nano-particles were first reported in a 2013 study by a team from the University of California (by Williams et al ) but these authors made a mistaken overestimation of their numbers based on the total number of “particles” between 50 and 1000 nano-meters, which would mostly include liquid droplets of the aerosol.

An estimate of the numbers of metallic particles and their composition in terms of metallic oxides was made by CDC researchers Pappas et al (2020) for aerosols generated by various pod devices (Juul, Blu and Vuse Alto). The numbers reported were below 2000 nano-particles per puff of different metal oxides (their Table VI). However, these authors found several outliers with over 20,000 particles per puff only in pods with certain e-liquid flavors (not in same pods with other e-liquid flavors).

Unfortunately, Pappas et al do not provide full information on the experimental data, only report absolute ranges of particle numbers without a statistical analysis (which renders the study unreproducible). This lack of information suggests that their outliers (also found in previous studies focused on metal element mass content) were most likely produced by defective or corroded cartridges or under overheating or dry puff conditions.

Nevertheless, even if we take their results at face value and considering that typical puffs in pod devices involve over 100 million liquid droplets, 2000 nano-particles (even 20,000 for the the outliers) form a negligible part of the particulate phase of vape aerosols (one metal particle in every 1-10 million droplets). This negligible contribution of metallic particles is reflected in well designed emission studies that quantified metallic element mass content under the Optimal Regime (see discussion further ahead).

Figure 3: one solid metal particle in every 10 million liquid particles

An article by Al-Qaysi and Abdulla published in 2021 reviews emission studies and analytic techniques focused on detection of metal contents in emission studies. They highlight 4 techniques, such ass inductively coupled plasma-mass spectrometry (ICP-MS), all based on ionized plasmas (detaching atoms into ions either heating the sample or using wavelengths in light beams as proxies for concentrations).

Al-Qaysi and Abdulla also reviewed Pappas et al and concluded that either nano-particle numbers or mass concentrations of metallic elements in vape aerosols do not pose serious toxicological risks. However, the literature contains several uncritical reviews of metal studies that emit alarmist conclusions from studies detecting high metallic content (nickel and lead). As Soulet and I showed in our 2022 review these outcomes were produced only in aerosols under overheating conditions.

Metal studies

The 2013 study by Williams et al mentioned before voiced alarmist conclusions on the yields of various metal elements that were found in the nano-particles. However, Farsalinos and Voudris showed that all these yields were orders of magnitude below toxicological markers: chronic Permissible Daily Exposure (PDE) from inhalation medications defined by the U.S. Pharmacopeia (cadmium, chromium, copper, lead and nickel), the Minimal Risk Level (MRL) of the Agency for Toxic Substances and Disease Registry (manganese) and the Recommended Exposure Limit (REL) defined by the National Institute of Occupational Safety and Health (aluminum, barium, iron, tin, titanium, zinc and zirconium).NIOSH. Williams et al mentioned that nickel levels in the vape aerosols were up to 200 times above levels in a Marlborough cigarette, but 200 times a negligible quantity is still negligible.

A more recent study by Olmedo et al in 2018, a team from Johns Hopkins School of Medicine, tested 54 devices and also voiced alarm claiming that their detected metal concentrations “exceeded current health-based limits (daily limit of the Agency for Toxic Substances Disease Registry, ATSDR) in close to 50% or more of the samples for Cr, Mn, Ni, and Pb”. This study had several flaws: it is unreproducible since the devices were not identified, but most significantly the authors miscalculated respiratory concentrations (computed for the puff volume, not for the tidal volume with which users actually inhale). Hence, as shown by Farsalinos and Rodu all metal concentrations they found were well below the ATSDR limit. In our 2022 review Soulet and I confirmed the arguments of Farsalinos and Rodu.

As high powered third generation sub-ohm devices became available and popular around 2016-2018, several emission studies testing these devices reported high levels of byproducts, not only organic (aldehydes and even carbon monoxide and free radicals), but also metals, specially nickel, copper, lead and chromium. The main reason for these alarming experimental outputs was puffing these sub-ohm devices at high powers above the Optimal Regime (see Post 3) with a low airflow rate: the CORESTA airflow of 1.1 L/min and small variations of it.

A study by Zhao et al of same research group as Olmedo et al from Johns Hopkins School of Medicine published a particularly misleading study, testing two high powered sub-ohm tank devices with a CORESTA airflow, an Ishtick 25 (Eleaf electronics 0.2 Ohms) and a SMOK (Smoketech 0.6 Ohms). The SMOK was tested at powers up 220 W, when the recommended power for the device was 40-80 W. Not surprisingly, the authors report enormous concerning concentrations of nickel, lead and other metals in the aerosol. However, these concentrations do not represent real life usage, they occurred because the authors generated an aerosol from this device under extreme overheating conditions.

Unfortunately, the authors of this and other studies (funded by public health institutions) do not realize that they are generating overheated and toxin laden aerosols by puffing high powered devices with an insufficient airflow. The authors claim to be just following the accepted CORESTA standard, as if this standard was universally applicable to all vape devices (this is not the case, as I argued in Post 4). They mention that they are testing high powered devices roughly as vapers puff them normally, claiming “normal” conditions just because the tanks are not depleted of e-liquid and thus dry puffs were avoided.

These experimental flaws show how disconnected and unaware these authors have been of consumer usage and how uninterested they are on verifying if actual users of sub-ohm devices puffed them with airflow rates around 1 L/min. The fact that users of these devices puff them with high airflow rates (the ‘direct to lung’ style) was well known at the time and available in vaping forums and magazines, but these authors ignored these sources for not being “peer reviewed”. As I showed in Post 4, there are physical reasons that justify the need for high airflow rates when puffing high powered devices.

The disconnection between laboratory studies and consumer habits can be remedied simply by incorporating vaping volunteers in the study protocol. Some studies have done so, but most emission studies testing high powered devices have kept (and still keep) puffing them with CORESTA or “CORESTA-like” airflow without involving human vapers. This problem does not occur in laboratory testing of low powered devices that are compatible with CORESTA airflows. It is also absent in industry funded studies, since all devices manufactured by the tobacco industry are low powered.

Our extensive literature review

In 2022 Sebastien Soulet and myself published two extensive literature reviews on emission studies, one focused on metals (12 studies), the other on organic byproducts (36 studies), all published between 2017 and 2022, mostly funded by public health institutions and universities, some also from authors employed by industry.

This is a summary of the review. We examined the experimental quality of the studies in terms of 3 conditions, divided in 5 items:

1. Experimental consistency in setting up puffing protocols (vaping regime)

2. Experimental consistency in setting up the Power range,

3. Reproducibility of aerosol generation

4. Reproducibility of analytic methods

5. Toxicological consistency (correct exposure computation)

We graded them with tick marks in three levels (approval, failure and partial aproval/failure, see table in Figure 4). The final grade is given in a traffic light system to grade them as “reliable” (GREEN, at least 3 of the 4 conditions), “partially reliable” (to be taken with skepticism) (ORANGE, 2 conditions) and “unreliable” (RED, zero or one condition). Of the 12 metal studies, only 3 were reliable (GREEN), 4 were partially reliable (ORANGE) and 5 were unrelaible (RED).

Figure 4. Summary of the review of 1 metal studies

The comments in the last column are:
(1) sub-ohm device with CORESTA at high powers (certain overheating),
(2) sub-ohm device with CORESTA recommended powers (likely overheating),
(3) other forms of inconsistent protocol,
(4) incorrect computation of exposure,
(5) outliers not properly identified,
(6) devices not fully identified (unreproducible),
(7) testing power not identified (unreproducible),
(8) too frequent puffs,
(9) too long puffs,
(10) used too old devices (corrosion)

At least 6 of the 12 studies (50%) did not provide sufficient information to reproduce the experiments. High metal levels were found in 4 studies (one third of them) that puffed high powered devices with CORESTA or CORESTA-like airflows. Four studies examined old devices likely subjected to corrosion. Only 3 studies were reliable and all found negligible levels of metal element yields: Beuval and Palazzolo examined 2nd generation devices and the Juul was examined by Chen.

An easily avoidable experimental flaw (number (1) above) comes from testing high powered devices with an insufficient airflow. Such aerosols would be repelling to consumers. This flaw is not restricted to emission studies focusing on metals, but also occurs in emission studies targeting organic byproducts (as shown in our review on organic byproducts) and in pre-clinical studies that expose cell lines and rodents to aerosols generated under these overheating conditions.

Unfortunately, there are several literature reviews on cancer risk from metals in vape aerosols that uncritically cite outcomes of emission studies that generated overheated aerosols (see Fowles et al). These risk models and the the emission studies they use as input data are often regarded (mistakenly) by health institutions and regulators as bona fide assessments of the safety of vape aerosols.

Conclusions & further posts.

As long as vape devices are puffed under the Optimal Regime (normal usage), the presence of metals in the aerosol is negligible. All studies reporting metal levels above toxicological safety markers were conducted under unrealistic and abnormal conditions that produce aerosols repellent to users. Alarming statements on health risks from metals in the aerosol are unwarranted.

In the following Substack posts I will discuss emission studies that have focused on organic byproducts (aldehydes, carbonyls, CO, free radicals). Sebastien Soulet and I published an extensive review of such studies. We have also identified 110 pre-clinical that expose cell lines and rodents to overheated and toxin laden aerosols generated by puffing sub-ohm devices with low airflow rates (which the authors regard as “normal” usage). A review of 40 of such studies is currently under peer review.