Vol.

7, 2022-10

Molecular Modeling of Macroscopic Phase Changes 1: Liquid Evaporation

Hugo Hernandez
ForsChem Research, 050030 Medellin, Colombia
[email protected]

doi: 10.13140/RG.2.2.29272.98566

Abstract

Phases and states of matter are macroscopic subjective notions about materials. At the
molecular scale, those notions become fuzzy or even non-existent. However, understanding
molecular motion and interaction helps describing the dynamics of phase changes at the
macroscopic scale. In this report, a simple molecular phase transfer model is presented, where
some molecules reach the interface after colliding with other molecules, and is able to cross
the boundary as long as its kinetic energy overcomes a certain barrier for transfer. The
efficiency of such transfer is expressed in terms of the error function of the energy barrier. This
model can be used to describe the evaporation of liquids. In this case, a general mechanistic
model for describing the vapor pressure of a pure compound is obtained, which can replace
common empirical models including Antoine equation. This mechanistic model requires only
two parameters, both of them having a clear physical meaning. Different compounds, and
particularly water, are used to show the consistent of this new model with currently accepted
empirical vapor pressure models.

Keywords

Antoine Equation, Energy Barrier, Evaporation, Flux, Interface, Phase Change, States of Matter,
Vapor Pressure

1. Introduction

The concept of phase in physics and chemistry is commonly interpreted as a state of matter.
However, phase and state of matter are actually different concepts. In fact, different phases in
a system may be found in the same state of matter, like for example liquid oil/liquid water
systems. Before discussing phase changes, let us clarify the basic concepts being used in this
report.

Cite as: Hernandez, H. (2022). Molecular Modeling of Macroscopic Phase Changes 1: Liquid Evaporation.
ForsChem Research Reports, 7, 2022-10, 1 - 20. Publication Date: 15/07/2022.
 Molecular Modeling of Macroscopic
Phase Changes 1: Liquid Evaporation
Hugo Hernandez
ForsChem Research
[email protected]

States of matter: The state of a material system is a categorical classification based on the
observable macroscopic properties of the system [1]. Thus, we may have a gas for compressible
materials with no fixed volume or shape, a liquid for (almost) incompressible materials with no
fixed shape, and a solid for incompressible materials with definite shape and volume. While
these are the three basic states of matter, some additional states can be defined based on
other properties, such as electrical conductivity (e.g. plasma, superconductor, etc.) and
viscosity (e.g. superfluid, glass, etc.), or they can be defined as intermediate states (e.g. liquid
crystals, supercritical fluids), or mixtures of materials microscopically present at different states
(i.e. colloids). These categorical states of matter are rather subjective, and depend on the
particular macroscopic properties considered. Therefore, many possible categories and
classifications are, in principle, possible.

Phase: A phase is a spatially-continuous region composed of molecules with an almost zero

time-average net external force, relative to the system [2]. This definition does not consider the
particular state of matter of the region considered or any adjacent region. The spatial
continuity requires a phase to always be enclosed by an interface. The almost zero time-
average net force implies that a phase needs to be stable long enough to be observed
macroscopically. Of course, this introduces some difficulties when considering systems out of
equilibrium. In an extreme scenario, each individual molecule represents a different phase until
macroscopically stable phases are obtained. The definition of a phase is also subjective, as it
depends on what is considered an almost zero time-average net force.

Interface (or interphase): Boundary between two adjacent or neighbor phases. These
boundaries are characterized by a drastic change in one or more properties of the system, and
therefore, by non-zero time-average net external forces acting on the molecules located at the
interface [3]. Precisely those forces give rise to interfacial tension. The above-mentioned
drastic change in macroscopic properties is also a subjective notion.

Since the properties of different states of matter are different, an interface will always exist
between different adjacent states. Therefore, each state will exist as a different phase of the
system. However, an interface can also be present between similar states of different materials
(e.g. liquid-liquid or solid-solid), indicating that different states represent different phases, but
different phases do not imply different states§.

It is also possible to find many different phases within the same state of matter. This is the
case, for example, of colloids and other segregated systems (not necessarily microscopically
segregated like colloids). This also occurs when drastic changes in local temperatures,
pressures, compositions, configurations, etc. are present in the system, even if the state of

§
Notice that not all properties of matter change between different states, but at least one arbitrary
property changes between different phases.

15/07/2022 ForsChem Research Reports Vol. 7, 2022-10 (2 / 20)

www.forschem.org
 Molecular Modeling of Macroscopic
Phase Changes 1: Liquid Evaporation
Hugo Hernandez
ForsChem Research
[email protected]

matter is the same and a clear interface is not observed. Consider for example the mirage
effect caused by a drastic difference in refractive index (due to drastic changes in temperature)
of air close to a hot surface. Segregated phases having similar properties are in fact, different
phases. For modeling purposes they might be represented for simplicity by a single phase.

Notice that all these definitions are subjective. The reason for this is that they originate from
the qualitative, macroscopic observation of the system. For example, clear differences are
easily observed in the behavior of gases and liquids. However, at higher pressures and
temperatures (supercritical conditions) those differences are no longer evident. Despite the
fact that these concepts are defined exclusively at a macroscopic scale, phase change (or phase
transfer) phenomena are ultimately caused by molecular motion and interaction. However, at
the molecular scale the concepts of states, phases and interfaces are even fuzzier. It is not
possible to determine the state of a particular molecule without observing other molecules in a
relatively large neighborhood. Whenever a molecule goes from one neighborhood to an
adjacent region with different macroscopic properties, we can say that the molecule has been
transferred from one phase to another. Such process necessarily implies that the molecule at
some point was part of the interface. The exact place and moment when a molecule leaves
phase and becomes part of the interface, or part of another phase, is impossible to determine
in practice [3]. This is because the concepts of phases and interfaces are defined at the
macroscopic scale and not at a molecular level.

New phases can also be created by a particular arrangement or aggregation of molecules,

resulting in regions with different properties. Such process occurs spontaneously as a result of
molecular collisions and interactions. Of course, not all collisions lead to new phase formation,
as certain geometrical and energetic conditions must be met. The newly formed phases can
also disappear as a result of molecular motion, when the particular arrangement of the new
phase changes into a new arrangement. Also, the newly formed phases can grow by molecular
transfer, since new interfaces are also created. One particular case of molecular re-
arrangement after collision is the chemical reaction, where the re-arrangement also involves
the relative position of atoms and/or subatomic entities (particularly electrons). The resulting
products of the reaction may affect the balance of forces, creating new phases or simply
promoting molecular transfer between phases.

Difficulties naturally arising from the use of macroscopic subjective notions make the modeling
and understanding of phase change processes at the molecular scale highly challenging. In
Section 2, a general model of macroscopic phase change by molecular transfer based on an
energy barrier is presented. Section 3 shows a particular application example in the case of
evaporation of a liquid.

15/07/2022 ForsChem Research Reports Vol. 7, 2022-10 (3 / 20)

www.forschem.org
 Molecular Modeling of Macroscopic
Phase Changes 1: Liquid Evaporation
Hugo Hernandez
ForsChem Research
[email protected]

2. Phase Transfer Model

For the case of phase change by molecular transfer, we will assume that an interface already
exist between two adjacent phases, as illustrated in Figure 1. The upper molecule is transferred
from phase 𝐴 to phase 𝐵, whereas the lower molecule is transferred from phase 𝐵 to phase 𝐴.

Figure 1. Molecular transfer across an interface between two adjacent phases.

Since the average net force acting on the molecules at the interface is not zero, a resistance to
transfer across the interface exists which can be represented as a probability of successful
transfer 𝑃(𝐴 → 𝐵) given as a function of an equivalent “energy barrier”.[3,4]

Assuming a normal distribution of random intermolecular forces and molecular velocities in

each direction, the probability of successful transfer, or transfer efficiency, can be determined
as the probability of a molecule moving in the positive direction subject to a net average force
in the opposite direction, relative to the probability under pure Brownian motion [4]:

𝐸𝐴→𝐵
𝜖𝐴→𝐵 = 1 − erf ( )
6𝑘𝐵 𝑇𝐴
(2.1)

where 𝐸𝐴→𝐵 ≥ 0 is the equivalent “energy barrier” for crossing the interface from phase 𝐴 to
phase 𝐵, 𝑘𝐵 = 1.38 × 10−23 𝐽/𝐾 is Boltzmann’s constant, 𝑇𝐴 is the temperature of phase 𝐴,
and erf represents the error function. Such equivalent energy barrier is not necessarily
constant, as it may change with the environmental conditions (e.g. pressure, temperature, etc.)

In general, we may say that

𝐸𝐴→𝐵 ∙ 𝐸𝐵→𝐴 ≥ 0
(2.2)

15/07/2022 ForsChem Research Reports Vol. 7, 2022-10 (4 / 20)

www.forschem.org
 Molecular Modeling of Macroscopic
Phase Changes 1: Liquid Evaporation
Hugo Hernandez
ForsChem Research
[email protected]

indicating that only positive energy barriers can be observed, and that they can be different.

At any given moment, the flux 𝐹𝑗𝐴→𝐵 of molecules of a particular chemical species 𝑗 from phase
𝐴 to phase 𝐵 across an interface of area 𝑆𝐴/𝐵 can be expressed as:

ℛ𝑗𝐴→𝐵
𝐹𝑗𝐴→𝐵 = 𝜖𝐴→𝐵
𝑆𝐴/𝐵
(2.3)

where ℛ𝑗𝐴→𝐵 is the rate of collision against the interface between phases 𝐴 and 𝐵 of molecules
of species 𝑗 in phase 𝐴. Similarly, the flux from phase 𝐵 to phase 𝐴 is given by:

ℛ𝑗𝐵→𝐴
𝐹𝑗𝐵→𝐴 = 𝜖
𝑆𝐴/𝐵 𝐵→𝐴
(2.4)
For ideal gases, the molecular fluxes become [5]:
2
𝐴→𝐵 𝐴→𝐵
𝜗𝑗⊥ (𝜗𝑗⊥ )
𝑘𝐵 𝑇𝐴 𝐴 𝜋 𝐴→𝐵 −
𝐹𝑗𝐴→𝐵 = 𝜖𝐴→𝐵 √ 𝜌̃𝑗 [√ 𝜗𝑗⊥ (erf ( ) + 1) + 𝑒 2 ]
2𝜋𝑚𝑗 2 √2
(2.5)
2
𝐵→𝐴 𝐵→𝐴
𝜗𝑗⊥ (𝜗𝑗⊥ )
𝑘𝐵 𝑇𝐵 𝐵 𝜋 𝐵→𝐴
𝐹𝑗𝐵→𝐴 = 𝜖𝐵→𝐴 √ 𝜌̃𝑗 [√ 𝜗𝑗⊥ (erf ( ) − 1) + 𝑒 − 2 ]
2𝜋𝑚𝑗 2 √2
(2.6)

where 𝑇 is the temperature at each phase, 𝑚𝑗 is the molecular mass of species 𝑗, 𝜌̃𝑗 is the
molecular density of species 𝑗 at each phase, and

𝑚𝑗
𝜗𝑗⊥ = √ 〈𝑣 〉
𝑘𝐵 𝑇 ⊥
(2.7)

is a dimensionless velocity given by the ratio between the average relative velocity of the
system perpendicular to the surface (〈𝑣⊥ 〉) and the standard deviation in thermal speed of the
system, also perpendicular to the surface.

Since molecules are permanently moving in both directions, the net flux from phase 𝐴 to phase
𝐵 across the interface is:

15/07/2022 ForsChem Research Reports Vol. 7, 2022-10 (5 / 20)

www.forschem.org
 Molecular Modeling of Macroscopic
Phase Changes 1: Liquid Evaporation
Hugo Hernandez
ForsChem Research
[email protected]

𝐴→𝐵
ℛ𝑗𝐴→𝐵 𝜖𝐴→𝐵 − ℛ𝑗𝐵→𝐴 𝜖𝐵→𝐴
𝐹𝑗,𝑛𝑒𝑡 = 𝐹𝑗𝐴→𝐵 − 𝐹𝑗𝐵→𝐴 = 𝐵→𝐴
= −𝐹𝑗,𝑛𝑒𝑡
𝑆𝐴/𝐵
(2.8)

In general, assuming a homogeneous composition of each phase, we might say that:

ℛ𝑗𝐴→𝐵 = 𝜅𝐴 𝜌̃𝑗𝐴 𝑉𝐴
(2.9)
𝐴
where 𝑉 is the volume of phase 𝐴, and 𝜅𝐴 is a proportionality factor, and therefore:

𝐴→𝐵
𝜅𝐴 𝜌̃𝑗𝐴 𝑉𝐴 𝜖𝐴→𝐵 − 𝜅𝐵 𝜌̃𝑗𝐵 𝑉𝐵 𝜖𝐵→𝐴
𝐹𝑗,𝑛𝑒𝑡 =
𝑆𝐴/𝐵
(2.10)
At equilibrium, when the net flux of molecules across the interface is zero, we have:

𝜅𝐴 𝜌̃𝑗𝐴 𝑉𝐴 𝜖𝐴→𝐵 = 𝜅𝐵 𝜌̃𝑗𝐵 𝑉𝐵 𝜖𝐵→𝐴

(2.11)

Since the fraction of molecules of species 𝑗 present only in phase 𝐴 with respect to the
molecules of species 𝑗 present in both phases is:

𝜌̃𝑗𝐴 𝑉𝐴
𝜑𝑗𝐴 =
𝜌̃𝑗𝐴 𝑉𝐴 + 𝜌̃𝑗𝐵 𝑉𝐵
(2.12)
Then Eq. (2.11) can be restated as follows:

𝜑𝑗𝐴 𝜅𝐴 𝜖𝐴→𝐵 = (1 − 𝜑𝑗𝐴 )𝜅𝐵 𝜖𝐵→𝐴

(2.13)

Solving for 𝜑𝑗𝐴 we obtain at equilibrium:

𝜖𝐵→𝐴
𝜑𝑗𝐴 =
(𝜅𝐴 /𝜅𝐵 )𝜖𝐴→𝐵 + 𝜖𝐵→𝐴
(2.14)
and equivalently,

(𝜅𝐴 /𝜅𝐵 )𝜖𝐴→𝐵

𝜑𝑗𝐵 = = 1 − 𝜑𝑗𝐴
(𝜅𝐴 /𝜅𝐵 )𝜖𝐴→𝐵 + 𝜖𝐵→𝐴
(2.15)

Considering Eq. (2.1) we obtain:

15/07/2022 ForsChem Research Reports Vol. 7, 2022-10 (6 / 20)

www.forschem.org
 Molecular Modeling of Macroscopic
Phase Changes 1: Liquid Evaporation
Hugo Hernandez
ForsChem Research
[email protected]

𝐸
(1 − erf ( 𝐵→𝐴 ))
6𝑘𝐵 𝑇𝐵
𝜑𝑗𝐴 = = 1 − 𝜑𝑗𝐵
𝐸𝐴→𝐵 𝐸𝐵→𝐴
(𝜅𝐴 /𝜅𝐵 ) (1 − erf ( )) + (1 − erf ( ))
6𝑘𝐵 𝑇𝐴 6𝑘𝐵 𝑇𝐵
(2.16)
In the absence of energy barriers:
1
𝜑𝑗𝐴 = = 1 − 𝜑𝑗𝐵
(𝜅𝐴 /𝜅𝐵 ) + 1
(2.17)
Particularly, assuming 𝜅𝐴 /𝜅𝐵 = 1, the molecular fractions simply become 𝜑𝑗𝐴 = 𝜑𝑗𝐵 = 1/2.

If the energy barrier for transfer from phase 𝐴 to phase 𝐵 is much larger than the energy
barrier for the reverse transfer, then 𝜑𝑗𝐴 → 1 and 𝜑𝑗𝐵 → 0. Similarly, if the energy barrier for
transfer from phase 𝐵 to phase 𝐴 is much larger than the reverse barrier, then 𝜑𝑗𝐴 → 0 and
𝜑𝑗𝐵 → 1. The transfer efficiencies and molecular fractions as functions of the energy barrier
𝐸𝐴→𝐵 for 𝐸𝐵→𝐴 = 0 and 𝜅𝐴 /𝜅𝐵 = 1 are illustrated in Figure 2.

𝐸𝐴→𝐵
Figure 2. Effect of energy barrier 𝑘𝐵 𝑇
on the transfer efficiencies (left plot, Eq. 2.1) and
equilibrium molecular fractions (right plot, Eq. 2.16), for 𝐸𝐵→𝐴 = 0 and 𝜅𝐴 /𝜅𝐵 = 1.

The efficiency given in Eq. (2.1) can also be approximated by an exponential function, as
follows:
𝐸
− 𝐴→𝐵
𝜖𝐴→𝐵 ≈ 𝑒 3𝑘𝐵 𝑇𝐴

(2.18)
and therefore Eq. (2.14) and (2.15) become:
𝐸
− 𝐵→𝐴
𝑒 3𝑘𝐵 𝑇𝐵
𝜑𝑗𝐴 = 𝐸 𝐸
− 𝐴→𝐵 − 𝐵→𝐴
(𝜅𝐴 /𝜅𝐵 )𝑒 3𝑘𝐵 𝑇𝐴 +𝑒 3𝑘𝐵 𝑇𝐵

(2.19)

15/07/2022 ForsChem Research Reports Vol. 7, 2022-10 (7 / 20)

www.forschem.org
 Molecular Modeling of Macroscopic
Phase Changes 1: Liquid Evaporation
Hugo Hernandez
ForsChem Research
[email protected]

𝐸
− 𝐴→𝐵
(𝜅𝐴 /𝜅𝐵 )𝑒 3𝑘𝐵 𝑇𝐵
𝜑𝑗𝐵 = 𝐸 𝐸 = 1 − 𝜑𝑗𝐴
− 𝐴→𝐵 − 𝐵→𝐴
(𝜅𝐴 /𝜅𝐵 )𝑒 3𝑘𝐵 𝑇𝐴 + 𝑒 3𝑘𝐵𝑇𝐵
(2.20)

The corresponding transfer probabilities and equilibrium molecular fractions obtained for the
exponential approximation (for 𝐸𝐵→𝐴 = 0, and 𝜅𝐴 /𝜅𝐵 = 1) are presented in Figure 3.

𝐸𝐴→𝐵
Figure 3. Effect of energy barrier 𝑘𝐵 𝑇
on the transfer efficiencies (left plot, Eq. 2.18) and
equilibrium molecular fractions (right plot, Eq. 2.19 and 2.20) using an exponential
approximation, for 𝐸𝐵→𝐴 = 0, 𝑇𝐴 = 𝑇, and 𝜅𝐴 /𝜅𝐵 = 1. Dashed lines using the error function
model are included for comparison.

In the previous figures it was assumed that 𝜅𝐴 /𝜅𝐵 = 1, only for illustrative purposes. In
practice, the ratio 𝜅𝐴 /𝜅𝐵 might not only be different from 1, but also can be a function of
temperature, volume of each phase, and other variables. In addition, before reaching
equilibrium, as the net transfer of molecules from one phase to another is different from zero,
the volume of each phase may change, as well as the interfacial area and eventually the energy
barrier, increasing the complexity of this phenomenon.

3. Liquid Evaporation Model

As an example of phase change by molecular transfer, let us consider a model for the
evaporation of liquids. The system considered is depicted in Figure 4. It consists of a rigid
cuboid containing a pure substance in two different phases: Liquid and Gas (Vapor). The area of
the interphase between the liquid and gas phases is constant (𝑆𝐿/𝐺 ), corresponding to the
cross-section of the container.

15/07/2022 ForsChem Research Reports Vol. 7, 2022-10 (8 / 20)

www.forschem.org
 Molecular Modeling of Macroscopic
Phase Changes 1: Liquid Evaporation
Hugo Hernandez
ForsChem Research
[email protected]

Figure 4. Liquid evaporation in a closed container.

The molecular flux from the gas to the liquid phase can be modeled assuming ideal gas
behavior (from Eq. 2.5):
𝐺→𝐿 2
𝑘𝐵 𝑇𝐺 𝐺 𝜋 𝐺→𝐿 𝜗⊥𝐺→𝐿 (𝜗⊥ )
𝐹 𝐺→𝐿
= 𝜖𝐺→𝐿 √ 𝜌̃ [√ 𝜗⊥ (erf ( ) + 1) + 𝑒 − 2 ]
2𝜋𝑚 2 √2
(3.1)
where (assuming ideal gas behavior):

𝑁𝐺 𝑃𝐺
𝜌̃𝐺 = =
𝑉𝐺 𝑘𝐵 𝑇𝐺
(3.2)
𝑃𝐺 is the pressure exerted by the gas on the container, and (from Eq. 2.7):

𝑚 𝑚 1 𝑑𝑉𝐺
𝜗⊥𝐺→𝐿 = √ 〈𝑣⊥ 〉 = √ ( )
𝑘𝐵 𝑇𝐺 𝑘𝐵 𝑇𝐺 2𝑆𝐿/𝐺 𝑑𝑡
(3.3)

1 𝑑𝑉𝐺 𝑘𝐵 𝑇𝐺
Since 2𝑆𝐿/𝐺 𝑑𝑡
≪√ 𝑚
, we can safely assume that 𝜗⊥𝐺→𝐿 → 0, and therefore, Eq. (3.1)
becomes:
𝑃𝐺 𝜖𝐺→𝐿
𝐹 𝐺→𝐿 =
√2𝜋𝑚𝑘𝐵 𝑇𝐺
(3.4)
The maximum flux obtained across the interface is (for 𝜖𝐺→𝐿 = 1) [6]:

𝐺→𝐿
𝑃𝐺
𝐹𝑚𝑎𝑥 =
√2𝜋𝑚𝑘𝐵 𝑇𝐺
(3.5)

15/07/2022 ForsChem Research Reports Vol. 7, 2022-10 (9 / 20)

www.forschem.org
 Molecular Modeling of Macroscopic
Phase Changes 1: Liquid Evaporation
Hugo Hernandez
ForsChem Research
[email protected]

On the other hand, the molecular flux from the liquid to the gas phase is given by Eq. (2.3) and
Eq. (2.9):
𝜅𝐿 𝑉𝐿 𝐿
𝐹 𝐿→𝐺 = 𝜖𝐿→𝐺 𝜌̃
𝑆𝐿/𝐺
(3.6)

In the absence of molecular collisions and energy barriers (and assuming 𝜗⊥𝐿→𝐺 → 0), the
maximum molecular flux from the liquid would be:

𝑘𝐵 𝑇𝐿 𝐿
𝐿→𝐺
𝐹𝑚𝑎𝑥 =√ 𝜌̃
2𝜋𝑚
(3.7)
Thus, we may say that:

𝐹 𝐿→𝐺 = 𝜖𝐿→𝐺 𝜓𝑐𝑟 𝐹𝑚𝑎𝑥

𝐿→𝐺

(3.8)

𝜅𝐿 𝑉𝐿 2𝜋𝑚
where 𝜓𝑐𝑟 = √ is a crowding factor reflecting the effect of molecular collisions and
𝑆𝐿/𝐺 𝑘𝐵 𝑇𝐿

diffusion due to molecular crowding on the molecular flux, which is a function of the
temperature-dependent diffusion coefficient in the liquid. Thus,

𝑘𝐵 𝑇𝐿 𝐿
𝐹 𝐿→𝐺 = 𝜖𝐿→𝐺 𝜓𝑐𝑟 √ 𝜌̃
2𝜋𝑚
(3.9)

At chemical equilibrium, the molecular fluxes in both directions are identical, resulting in:

𝑒𝑞
𝑃𝐺 𝜖𝐺→𝐿 𝑘𝐵 𝑇𝐿 𝐿
= 𝜖𝐿→𝐺 𝜓𝑐𝑟 √ 𝜌̃
√2𝜋𝑚𝑘𝐵 𝑇𝐺 2𝜋𝑚
(3.10)
𝑒𝑞
where 𝑃𝐺 is the pressure exerted by the gas phase on the walls of the container at chemical
equilibrium. This equilibrium pressure will be denoted as the vapor pressure of the liquid 𝑃𝑣 :

𝑒𝑞 𝜖𝐿→𝐺
𝑃𝑣 = 𝑃𝐺 = 𝜓𝑐𝑟 𝑘 √𝑇 𝑇 𝜌̃𝐿
𝜖𝐺→𝐿 𝐵 𝐺 𝐿
(3.11)
Also assuming thermal equilibrium in the system:

15/07/2022 ForsChem Research Reports Vol. 7, 2022-10 (10 / 20)

www.forschem.org
 Molecular Modeling of Macroscopic
Phase Changes 1: Liquid Evaporation
Hugo Hernandez
ForsChem Research
[email protected]

𝜖𝐿→𝐺
𝑃𝑣 = 𝜓𝑐𝑟 𝑘 𝑇𝜌̃𝐿
𝜖𝐺→𝐿 𝐵
(3.12)

Notice that the terms 𝜓𝑐𝑟 , 𝜖𝐿→𝐺 , 𝜖𝐺→𝐿 and 𝜌̃𝐿 are functions of the system temperature. The
product 𝜓𝑐𝑟 𝜌̃𝐿 can be interpreted as an effective molecular liquid density at the interface.
Particularly, the energy barrier efficiencies to transfer can be expressed using Eq. (2.1) resulting
in:
𝐸
1 − erf ( 𝐿→𝐺 )
6𝑘𝐵 𝑇
𝑃𝑣 = 𝜓𝑐𝑟 ( ) 𝑘𝐵 𝑇𝜌̃𝐿
𝐸𝐺→𝐿
1 − erf ( )
6𝑘𝐵 𝑇
(3.13)

Assuming that 𝐸𝐺→𝐿 → 0 (as the gas does not offer a significant resistance to phase transfer)
and neglecting the effect of temperature on 𝜓𝑐𝑟 and 𝜌̃𝐿 we obtain:

𝐸𝐿→𝐺
𝑃𝑣 ≈ 𝜓𝑐𝑟 𝑘𝐵 𝜌̃𝐿 (1 − erf ( )) 𝑇
6𝑘𝐵 𝑇
(3.14)
The error function term can be approximated by an exponential term as follows:
𝐸
− 𝐿→𝐺
𝑃𝑣 ≈ 𝜓𝑐𝑟 𝑘𝐵 𝜌̃𝐿 𝑇𝑒 3𝑘𝐵 𝑇

(3.15)
and therefore:
𝐸𝐿→𝐺
ln 𝑃𝑣 ≈ ln(𝜓𝑐𝑟 𝑘𝐵 𝜌̃𝐿 ) + ln 𝑇 −
3𝑘𝐵 𝑇
(3.16)

Now, for a limited range of temperature values, Eq. (3.16) can be further approximated by any
of the following empirical expressions [7]:

 Dalton Equation:
ln 𝑃𝑣 ≈ 𝐴 + 𝐵𝑇
(3.17)
 August Equation:
𝐵
ln 𝑃𝑣 ≈ 𝐴 −
𝑇
(3.18)
 Antoine Equation:
𝐵
ln 𝑃𝑣 ≈ 𝐴 −
𝑇+𝐶
(3.19)

15/07/2022 ForsChem Research Reports Vol. 7, 2022-10 (11 / 20)

www.forschem.org
 Molecular Modeling of Macroscopic
Phase Changes 1: Liquid Evaporation
Hugo Hernandez
ForsChem Research
[email protected]

 Henglein Equation:
𝐵
ln 𝑃𝑣 ≈ 𝐴 −
𝑇𝑛
(3.20)
 Nernst Equation:
𝐶
ln 𝑃𝑣 ≈ 𝐴 + 𝐵 ln 𝑇 + − 𝐷𝑇
𝑇
(3.21)
 Bose Equation:
𝐵 𝐶 𝐷
ln 𝑃𝑣 ≈ 𝐴 − + 2+ 3
𝑇 𝑇 𝑇
(3.22)
Or in general, by the series expansion:

ln 𝑃𝑣 ≈ 𝐴 + 𝐵𝑇 + 𝐶𝑇 2 + 𝐷𝑇 3 + ⋯
(3.23)

The last expression is particularly useful when changes in liquid density and crowding factor
with temperature are considered.

Among these empirical expressions, the one most commonly used is Antoine Equation (Eq.
3.19). Values of the three parameters in Antoine Equation can be found for many pure
substances in typical chemical compound databases**. Table 1 shows Antoine parameters
values for a sample of different compounds††.

This sample of compounds is used to illustrate the equivalence between the Antoine model and
Eq. (3.14). Notice that Eq. (3.14) is based on molecular first-principles, it only requires the
estimation of two parameters (𝜓𝑐𝑟 and 𝐸𝐿→𝐺 ), and both parameters have a physical meaning.
The parameter values obtained by minimizing the sum of square error between Eq. (3.14) and
Eq. (3.19) over the applicable range of temperatures are summarized in Table 2. The fit
between both models (𝑅 2 coefficient) is also included. The average fit observed was 99.99%,
with a minimum fit of 99.96% for hydrochloric acid. These values clearly show the equivalence
between the models.

The values for the crowding factor 𝜓𝑐𝑟 obtained after fitting Eq. (3.14) to Antoine model for
different compounds resulted in values ranging between 0.28 and 6.71. This parameter is
mainly influenced by the shape and size of the molecules [8], although intermolecular forces
may also play an important role.

**
For example: http://webbook.nist.gov/chemistry/
††
Retrieved from : https://vacuu-lan.com/vapor-pressure-estimates-antoine-equation/

15/07/2022 ForsChem Research Reports Vol. 7, 2022-10 (12 / 20)

www.forschem.org
 Molecular Modeling of Macroscopic
Phase Changes 1: Liquid Evaporation
Hugo Hernandez
ForsChem Research
[email protected]

Table 1. Antoine coefficients for a sample of pure substances‡‡

Chemical Compound A [log10 bar] B [K] C [K] Applicable Temp. Range [K]

Acetic Acid 4.68206 1642.540 -39.764 290.26 391.01

Acetic Anhydride 4.24655 1427.770 -75.113 335.99 412.57
Acetone 4.42448 1312.253 -32.445 259.16 507.60
Acetonitrile 4.27873 1355.374 -37.853 288.30 362.30
Benzene 4.01814 1203.835 -53.226 287.70 354.07
Carbon Tetrachloride 4.02291 1221.781 -45.739 293.03 350.86
Chlorobenzene 4.11083 1435.675 -55.124 335.19 404.88
Chloroform 4.20772 1233.129 -40.953 215.00 334.40
Cyclohexane 3.96988 1203.526 -50.287 293.06 354.73
Cyclopentane 4.00288 1119.208 -42.412 288.86 323.18
Dimethyl Acetamide 6.09451 2725.960 28.209 303.00 363.00
Dimethyl Formamide 3.93068 1337.716 -82.648 303.00 363.00
Dimethyl Sulfoxide 5.23039 2239.161 -29.215 293.00 323.00
Dioxane 4.58135 1570.093 -31.297 293.00 378.00
Ethanol 5.37229 1670.409 -40.191 273.00 351.70
Ethyl Acetate 4.22809 1245.702 -55.189 288.73 348.98
Ethyl Ether 4.02200 1062.640 -44.930 250.04 328.57
Ethylene Dichloride 4.58518 1521.789 -24.670 242.33 372.60
Ethylene Glycol 4.97012 1914.951 -84.996 323.00 473.00
Heptane 4.02832 1268.636 -56.199 299.07 372.43
Hydrochloric Acid 4.57389 868.358 1.754 188.30 309.40
iso-Butanol 4.43126 1236.991 -101.528 353.36 388.77
iso-Propanol 4.57795 1221.423 -87.474 395.10 508.24
Methanol 5.20409 1581.341 -33.500 288.10 356.83
Methylene Chloride 4.53691 1327.016 -20.474 233.00 313.00
n-Butanol 4.54607 1351.555 -93.340 295.80 391.00
n-Hexane 4.00266 1171.530 -48.784 286.18 342.69
n-Propanol 5.31384 1690.864 -51.804 292.40 370.50
Pentane 3.98920 1070.617 -40.454 268.80 341.37
Pyridine 4.16272 1371.358 -58.496 340.50 426.04
tert-Butanol 4.49774 1174.869 -93.920 312.66 355.56
Tetrahydrofuran 4.12118 1202.942 -46.818 296.29 372.80
Toluene 4.14157 1377.578 -50.507 273.00 323.00
Water 6.20963 2354.731 7.559 293.00 343.00

A decrease in the value of 𝜓𝑐𝑟 is observed in small molecules as a result of the steric hindrance
caused by molecular collisions. However, as the hindrance increases (with size and particularly
for branched molecules), the mobility of molecules at the interface is reduced towards the
liquid phase, but is not hindered towards the vapor phase, resulting in a larger flux of
molecules from the liquid to the gas phase.

‡‡
These coefficients were obtained using a modified form of Antoine equation with a decimal logarithm:
𝐵
log10 𝑃𝑣 [𝑏𝑎𝑟] = 𝐴 −
𝑇[𝐾] + 𝐶
The coefficients for the natural logarithm form of Antoine equation are obtained by using a conversion
factor of ln 10 ≈ 2.302585 for the values of 𝐴 and 𝐵. Coefficient 𝐶 remains unchanged.

15/07/2022 ForsChem Research Reports Vol. 7, 2022-10 (13 / 20)

www.forschem.org
 Molecular Modeling of Macroscopic
Phase Changes 1: Liquid Evaporation
Hugo Hernandez
ForsChem Research
[email protected]

Table 2. Properties of a sample of pure substances and their corresponding parameters 𝜓𝑐𝑟 and
𝐸𝐿→𝐺 estimated from the Antoine model
Molecular Liquid Crowding Energy barrier Model fit
Chemical Compound 2
mass [g/mol] density [g/L] factor 𝝍𝒄𝒓 𝑬𝑳→𝑮 [kJ/mol] R
Acetic Acid 60.05 1050 0.97672 43.0235 99.98958%
Acetic Acid Anhydride 102.09 1080 1.68500 45.6944 99.99768%
Acetone 58.08 784 2.67508 38.8795 99.98259%
Acetonitrile 41.05 786 0.63128 37.3797 99.99019%
Benzene 78.11 876 1.01795 36.9656 99.99551%
Carbon Tetrachloride 153.82 1590 0.97280 36.0913 99.99462%
Chlorobenzene 112.56 1110 1.02799 42.4457 99.99470%
Chloroform 119.38 1490 0.98788 35.0844 99.99275%
Cyclohexane 84.16 774 1.08722 36.5273 99.99523%
Cyclopentane 70.10 751 1.03423 33.3065 99.99740%
Dimethyl Acetamide 87.12 940 0.88669 44.8640 99.97470%
Dimethyl Formamide 73.09 944 0.78619 44.7804 99.99952%
Dimethyl Sulfoxide 78.13 1100 0.28029 46.9220 99.99549%
Dioxane 88.11 1030 1.19411 40.2255 99.98685%
Ethanol 46.07 789 3.16946 42.3508 99.99194%
Ethyl Acetate 88.11 902 1.59320 37.9540 99.99626%
Ethyl Ether 74.12 713 1.44023 32.5652 99.99342%
Ethylene Dichloride 98.96 1250 1.15718 38.3070 99.98515%
Ethylene Glycol 62.07 1110 2.49598 57.1434 99.99769%
Heptane 100.21 684 1.61645 38.9496 99.99535%
Hydrochloric Acid 36.46 1200 1.77848 22.4411 99.95715%
iso-Butanol 74.12 802 5.55433 46.5584 99.99993%
iso-Propanol 60.10 786 6.51058 44.4694 99.99821%
Methanol 32.04 792 1.94150 40.1199 99.98958%
Methylene Chloride 84.93 1330 0.84236 32.8909 99.98381%
n-Butanol 74.12 810 4.73926 47.3238 99.99983%
n-Hexane 86.18 655 1.43853 35.4935 99.99555%
n-Propanol 60.10 803 4.55393 45.2139 99.99431%
Pentane 72.15 626 1.49693 32.3146 99.99204%
Pyridine 79.10 982 1.22288 42.1844 99.99425%
tert-Butanol 74.12 781 6.70967 43.7454 99.99999%
Tetrahydrofuran 72.11 890 1.21622 36.3040 99.99294%
Toluene 92.14 867 0.74574 38.8762 99.99622%
Water 18.02 997 0.51278 42.9862 99.98269%

Only those molecules with kinetic energy large enough to overcome the energy barrier will
effectively change phase. The energy barrier values found by fitting Eq. (3.14) to Antoine model
for the compounds considered ranged between 22.4 and 57.1 𝑘𝐽/𝑚𝑜𝑙. These values are
strongly related to the interfacial tension, which in turn is determined by the intermolecular
forces between the molecules.

Figure 5 shows the graphical comparison between Eq. (3.14) and Antoine model for the
particular cases of carbon tetrachloride (largest molecule in the sample), dimethyl sulfoxide
(lowest crowding factor), ethylene glycol (highest energy barrier), hydrochloric acid (lowest

15/07/2022 ForsChem Research Reports Vol. 7, 2022-10 (14 / 20)

www.forschem.org
 Molecular Modeling of Macroscopic
Phase Changes 1: Liquid Evaporation
Hugo Hernandez
ForsChem Research
[email protected]

energy barrier), tert-butanol (highest crowding factor), and water (smallest molecule). As it can
be evidenced in these plots, both models are practically equivalent for the corresponding
applicable ranges of temperature values.

Figure 5. Representative examples of the fit between Antoine Equation and Eq. (3.14)

Antoine equation (Eq. 3.19) is valid only for a limited range of applicable temperatures. It
cannot be extrapolated. From a mathematical perspective, notice that for 𝑇 ≈ −𝐶, an
𝐵
indetermination in the vapor pressure emerges. Furthermore, as 𝑇 → 0, 𝑃𝑣 → 𝑒 𝐴−𝐶 ≫ 0, which
is a physical impossibility as the lack of molecular mobility at 𝑇 → 0 should result in zero vapor
pressure. In addition, as 𝑇 → ∞, 𝑃𝑣 → 𝑒 𝐴 = constant, which is not consistent with ideal gas
behavior expected at extremely large temperatures. These inconsistencies clearly demonstrate
the empirical nature of Antoine equation. On the other hand, Eq. (3.14) results in the following

15/07/2022 ForsChem Research Reports Vol. 7, 2022-10 (15 / 20)

www.forschem.org
 Molecular Modeling of Macroscopic
Phase Changes 1: Liquid Evaporation
Hugo Hernandez
ForsChem Research
[email protected]

physically-consistent limits: As 𝑇 → 0, 𝑃𝑣 → 0; and as 𝑇 → ∞, 𝑃𝑣 ∝ 𝑇. Of course, Eq. (3.14) is only

an approximation because the terms 𝜓𝑐𝑟 , 𝜌̃𝐿 , and 𝐸𝐿→𝐺 can also be temperature-dependent.

In order to illustrate the advantage of using Eq. (3.14) over Antoine equation (Eq. 3.19) let us
consider the vapor pressure of water over a wide range of temperatures from 256 𝐾 to 573 𝐾.
A single set of parameters of Antoine equation cannot fully describe such range of
temperatures. At least two sets of Antoine parameters are required [9,10]. The NIST webbook
reports 7 different sets of Antoine parameters for this range [9-13]. Some of these parameters
are valid in overlapping temperature ranges, thus, an average Antoine model is used to fit Eq.
(3.14). The results obtained are shown in Figure 6. The model obtained fitted the average
Antoine model with 𝑅 2 = 99.96% (in the original scale). The model parameters obtained were:
𝑘𝐽
𝜓𝑐𝑟 = 0.91729 and 𝐸𝐿→𝐺 = 44.0608 . In the logarithm scale plot it is possible to observe a
𝑚𝑜𝑙
discontinuity in the average Antoine model because the temperature range between 373 𝐾
and 379 𝐾 is not covered by any of the parameter sets considered.

Figure 6. Water vapor pressure between 256 𝐾 and 573 𝐾. Dashed blue line: Average Antoine
Model. Solid green line: Best fit using Eq. (3.14). Left plot: Original scale. Right plot: Logarithm
scale.

Table 3. Water vapor pressure model parameter values for different temperature ranges
Antoine Equation [13] Eq. (3.14)
Applicable Temp. A B C 𝑬𝑳→𝑮 Model Original data
𝝍𝒄𝒓
Range (K) [log10 bar] [K] [K] [kJ/mol] fit source
379 573 3.55959 643.748 -198.043 0.51278 42.9862 99.983% [10]
273 303 5.40221 1838.675 -31.737 0.30468 41.6160 99.996% [11]
304 333 5.20389 1733.926 -39.485 0.47322 42.7614 99.997% [11]
334 363 5.0768 1659.793 -45.854 0.66361 43.7828 99.998% [11]
344 373 5.08354 1663.125 -45.622 0.73259 44.1124 99.998% [11]
293 343 6.20963 2354.731 7.559 0.91378 44.0342 99.941% [12]
256 373 4.6543 1435.264 -64.848 0.58768 43.3672 99.997% [9]
Average
256 573 0.91729 44.0608 99.960%
model
Absolute coefficient
16.0% 31.8% 108.7% 32.8% 2.0%
of variation (%)

15/07/2022 ForsChem Research Reports Vol. 7, 2022-10 (16 / 20)

www.forschem.org
 Molecular Modeling of Macroscopic
Phase Changes 1: Liquid Evaporation
Hugo Hernandez
ForsChem Research
[email protected]

Table 3 shows the differences in parameter values obtained for different temperature ranges,
indicating that the parameters used by Eq. (3.14) are more robust than the parameters used by
Antoine equation, showing less relative variation in the parameter values, particularly in the
energy barrier term 𝐸𝐿→𝐺 with a coefficient of variation of only 2%, while the corresponding
energy term coefficient (𝐵) in Antoine equation has a coefficient of variation of 31.8%. In the
Antoine model, the parameter 𝐴 shows less variation (16%). However, if we consider it as a
pre-exponential term (10 𝐴 ), the observed coefficient of variation becomes actually 173%,
which can be compared with the parameter 𝜓𝑐𝑟 with 32.8% variation.

The most reliable set of parameters is that obtained for the broadest range of temperature
values, which in this case are obtained for the average model from 256 𝐾 to 573 𝐾. Of course,
the parameters will be more reliable if they are fitted directly from experimental data instead
of using the Antoine model. The parameters obtained using the original data vapor pressure
𝑘𝐽
reported for water from 256 𝐾 to 573 𝐾 [9,10], are 𝜓𝑐𝑟 = 1.25444 and 𝐸𝐿→𝐺 = 45.5056 𝑚𝑜𝑙.
The fit of this model was 𝑅 2 = 99.57% (in the both original and logarithm scales), as illustrated
in Figure 7.

Figure 7. Water vapor pressure between 256 𝐾 and 573 𝐾. Blue dots: Experimental data [9,10].
Solid green line: Best fit using Eq. (3.14). Left plot: Original scale. Right plot: Logarithm scale.

Finally, let us consider the equilibrium phase partition for this evaporation model. From Eq.
(2.12) we obtain:
𝜑𝐺 𝜌̃𝐿 𝑉𝐿 𝜑 𝐺 𝜌̃𝐿 (1 − 𝜙 𝐺 )
𝜌̃𝐺 = =
(1 − 𝜑𝐺 )𝑉𝐺 (1 − 𝜑𝐺 )𝜙 𝐺
(3.24)
𝐺 𝑉𝐺
where 𝜙 = is the volume fraction of gas phase in the system.
𝑉𝐺 +𝑉𝐿

For an open system where 𝑉𝐺 → ∞, we simply obtain 𝜙 𝐺 → 1, indicating that all the liquid will
eventually evaporate, even when the system temperature is below the boiling point.

15/07/2022 ForsChem Research Reports Vol. 7, 2022-10 (17 / 20)

www.forschem.org
 Molecular Modeling of Macroscopic
Phase Changes 1: Liquid Evaporation
Hugo Hernandez
ForsChem Research
[email protected]

For a closed system (constant number of molecules) with a constant volume, it can be shown§§
that:
𝜌̃𝐿 − 𝜌̃
𝜙𝐺 = 𝐿
𝜌̃ − 𝜌̃𝐺
(3.25)
𝑁𝐿 +𝑁𝐺
where 𝜌̃ = is the constant overall molecular density in the system.
𝑉𝐿 +𝑉𝐺

Thus, Eq. (3.24) can be expressed as follows (solving for 𝜑𝐺 ):

1
𝜑𝐺 =
𝜌̃
−1
𝜌̃𝐺
1+ 𝜌̃
1− 𝐿
𝜌̃
(3.26)
Now, assuming an ideal gas phase we have:
𝑃𝑣
𝜌̃𝐺 =
𝑘𝐵 𝑇
(3.27)
therefore,
1
𝜑𝐺 =
𝜌̃𝑘𝐵 𝑇
−1
𝑃
1 + 𝑣 𝜌̃
1− 𝐿
𝜌̃
(3.28)

and using Eq. (3.14):

1
𝜑𝐺 =
𝜌̃/𝜌̃𝐿
𝐸 −1
𝜓𝑐𝑟 (1 − erf ( 𝐿→𝐺 ))
6𝑘𝐵 𝑇
1+
1 − 𝜌̃/𝜌̃𝐿
(3.29)

Figure 8 illustrates the effect of the overall mass density and temperature of the system on the
equilibrium vapor fraction of acetone. Depending on the volume available for evaporation, and
on the system temperature, different equilibrium partitions between liquid and vapor are
obtained. For a constant temperature and initial amount of substance, a decrease in the
volume of the container (increase in the overall density) results in a higher fraction of liquid.

𝑁𝐿 𝑁𝐿 +𝑁𝐺 𝑁𝐿 (𝑉𝐿 +𝑉𝐺 )−(𝑁𝐿 +𝑁𝐺 )𝑉𝐿 𝑁𝐿 𝑉𝐿 +𝑁𝐿 𝑉𝐺 −𝑁𝐿 𝑉𝐿 −𝑁𝐺 𝑉𝐿

̃ 𝐿 −𝜌
̃ − 𝑉𝐺 (𝑁𝐿 𝑉𝐺−𝑁𝐺 𝑉𝐿 )
§§ 𝜌 𝑉𝐿 𝑉𝐿 +𝑉𝐺 𝑉𝐿 (𝑉𝐿 +𝑉𝐺 ) 𝑉𝐿 +𝑉𝐺 𝑉𝐺
̃ 𝐿 −𝜌
̃𝐺
= 𝑁𝐿 𝑁𝐺 = 𝑁𝐿 𝑉𝐺 −𝑁𝐺 𝑉𝐿 = 𝑁𝐿 𝑉𝐺 −𝑁𝐺 𝑉𝐿 = (𝑉 = = 𝜙𝐺
𝜌 − 𝐿 +𝑉𝐺 )(𝑁𝐿 𝑉𝐺 −𝑁𝐺 𝑉𝐿 ) 𝑉𝐿 +𝑉𝐺
𝑉𝐿 𝑉𝐺 𝑉𝐿 𝑉𝐺 𝑉𝐺

15/07/2022 ForsChem Research Reports Vol. 7, 2022-10 (18 / 20)

www.forschem.org
 Molecular Modeling of Macroscopic
Phase Changes 1: Liquid Evaporation
Hugo Hernandez
ForsChem Research
[email protected]

Also, considering the same overall density, and increase in temperature increases the vapor
fraction until all the substance has completely evaporated.

Figure 8. Equilibrium vapor fraction for acetone at different conditions.

4. Conclusion

All macroscopic phenomena can be explained by the motion and interaction of matter at the
molecular level. Phase change is not an exception. A general mechanistic model of phase
change is presented, based on the transfer of molecules across an interface. An energy barrier
to transfer emerges as a result of the imbalance of forces at the interface. This approach is
used to describe the evaporation of liquids, resulting in a vapor pressure model with two
parameters (the energy barrier and a crowding factor). This model can replace the empirical
Antoine model which uses three parameters with no clear physical meaning. Different
substances were considered in order to illustrate the general validity of the proposed model.

Acknowledgment and Disclaimer

This report provides data, information and conclusions obtained by the author(s) as a result of original
scientific research, based on the best scientific knowledge available to the author(s). The main purpose
of this publication is the open sharing of scientific knowledge. Any mistake, omission, error or inaccuracy
published, if any, is completely unintentional.

This research did not receive any specific grant from funding agencies in the public, commercial, or not-
for-profit sectors.

15/07/2022 ForsChem Research Reports Vol. 7, 2022-10 (19 / 20)

www.forschem.org
 Molecular Modeling of Macroscopic
Phase Changes 1: Liquid Evaporation
Hugo Hernandez
ForsChem Research
[email protected]

This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International (CC BY-NC
4.0). Anyone is free to share (copy and redistribute the material in any medium or format) or adapt
(remix, transform, and build upon the material) this work under the following terms:
 Attribution: Appropriate credit must be given, providing a link to the license, and indicating if
changes were made. This can be done in any reasonable manner, but not in any way that
suggests endorsement by the licensor.
 NonCommercial: This material may not be used for commercial purposes.

References

[1] Brown, T. L., LeMay, H. E., Bursten, B. E., & Burdge, J. R. (2003). Chemistry: The Central Science. 9th
Edition. Prentice Hall, Inc. Chapter 1. Section 1.2: Classifications of Matter. pp. 5-8.
https://www.pearson.com/us/higher-education/program/Brown-Chemistry-The-Central-Science-9th-
Edition/PGM98410.html.
[2] Hernandez, H., & Tauer, K. (2021). Heterophase Polymerization: Basic Concepts and Principles.
Jenny Stanford Publishing. Section 1.4: Molecular Phases. pp. 26-31. doi: 10.1201/9781003119296.
[3] Hernandez, H., & Tauer, K. (2021). Heterophase Polymerization: Basic Concepts and Principles.
Jenny Stanford Publishing. Section 1.5: Molecular Behavior at Interfaces. pp. 32-36. doi:
10.1201/9781003119296.
[4] Hernández, H. F. (2008). Multiscale Simulation of Heterophase Polymerization: Application to the
Synthesis of Multicomponent Colloidal Polymer Particles. Doctoral Dissertation. Universität
Potsdam, Potsdam. https://publishup.uni-potsdam.de/frontdoor/index/index/docId/2370.
[5] Hernandez, H. (2019). Calculation of Molecular Fluxes and Equivalent Pressure in Ideal Gases.
ForsChem Research Reports, 4, 2019-03, 1-31. doi: 10.13140/RG.2.2.35898.44483.
[6] Beverley, K., Clint, J., & Fletcher, P. I. (1999). Evaporation rates of pure liquids measured using a
gravimetric technique. Physical Chemistry Chemical Physics, 1 (1), 149-153. doi: 10.1039/A805344H.
[7] Thomson, G. W. (1946). The Antoine equation for vapor-pressure data. Chemical Reviews, 38 (1), 1-
39. doi: 10.1021/cr60119a001.
[8] Hernandez, H. F., & Tauer, K. (2007). Brownian Dynamics simulation studies on radical capture in
emulsion polymerization. Macromolecular Symposia, 259 (1), 274-283. doi: 10.1002/masy.200751332.
[9] Stull, D. R. (1947). Vapor pressure of pure substances. Organic and inorganic compounds. Industrial
& Engineering Chemistry, 39 (4), 517-540. doi: 10.1021/ie50448a022.
[10] Liu, C. T., & Lindsay Jr, W. T. (1970). Vapor pressure of deuterated water from 106 to 300. deg.
Journal of Chemical and Engineering Data, 15 (4), 510-513. doi: 10.1021/je60047a015.
[11] Bridgeman, O.C., & Aldrich, E.W. (1964). Vapor Pressure Tables for Water. Journal of Heat Transfer,
86 (2), 279-286. doi: 10.1115/1.3687121.
[12] Gubkov, A. N., Fermor, N. A., & Smirnov, N. I. (1964). Vapor pressure of mono-poly systems. Zh.
Prikl. Khim. (Leningrad), 37, 2204-2210.
[13] National Institute of Standards and Technology NIST (2021). NIST Chemistry WebBook, SRD 69:
Water. https://webbook.nist.gov/cgi/cbook.cgi?ID=C7732185&Units=SI&Mask=4#Thermo-Phase.
Retrieved: July 14, 2022.

15/07/2022 ForsChem Research Reports Vol. 7, 2022-10 (20 / 20)

www.forschem.org