Skip to main content
NCBI home page
ACS Omega logoLink to ACS Omega
. 2024 May 21;9(22):24039–24049. doi: 10.1021/acsomega.4c02606

Molecular Dynamics Simulation of Ignition Behavior of Low Palmitic Acid-Coated Nanoaluminum Powder

Yi Liu 1, Jiangtao Xu 1, Pingan Liu 1,*, Lei Wang 1, Dianlong Sun 1, Hancong Liu 1
PMCID: PMC11154943  PMID: 38854524

Abstract

graphic file with name ao4c02606_0017.jpg

Large-scale Atomic/Molecular Massively Parallel Simulator (LAMMPS) was used to perform molecular dynamics (MD) simulations of the phase transition and decoating behavior of aluminum nanopowder (ANP)-palmitate composite particles under typical water ram engine conditions. We originally intended to investigate the effect of the degree of coating on the decorrelation behavior of the composite particles but accidentally discovered the premixed ignition behavior of low-coated composite particles. Therefore, we summarized and subdivided the four stages of precombustion adsorption, premixed ignition, melt-off, and full-scale combustion of palmitic acid-coated nanoaluminum powders by combining the simulations and studies of palmitic acid pyrolysis, ANP phase transition, and water molecule adsorption efficiency. We unexpectedly found that among the influencing factors of premixed ignition, the influence of hot and cold mixing degrees was greater than that of the ignition temperature.

1. Introduction

Aluminum-based propellant additives have received increasing attention due to their excellent energetic properties.1−3 Conventional aluminum powder additives suffer from combustion stability deficiencies such as ignition temperature and burning rate, which can be greatly improved by the application of aluminum nanopowder (ANP).3,4 However, due to the amazing specific surface area of ANP, its oxidative loss during storage increases significantly,5,6 which makes ANP coating technology a hot research topic.

Among the existing coating schemes, organic coatings have highly energetic properties and have attracted much attention and research.7,8 The properties of palmitic acids, such as melting point, relative molecular weight, and polarity (water resistance), satisfy many of the requirements of ANP coating materials9,10 and become the simulated coating material in this paper.

We know that the melting behavior of ANP takes precedence over the combustion and oxidation behavior during ignition,11,12 and the melting behavior of ANP directly destroys the adhesion of palmitic acid molecules, so the ignition behavior of palmitic acid compounds ANP tends to occur after the coating is detached.

Efficient hot and cold mixing is the basis for stable engine combustion. This water vapor with a large temperature span is more common in structures such as swirl combustion chambers,13 sudden broad combustion chambers,14 and blunt body flame stabilizers.15

Conventional molecular simulations of metal particle combustion do not take into account the premixing of the workmass.16,17 When ANP is heated by external water molecules, the temperature uniformity of these water molecules depends on the degree of hot and cold mixing of water vapor. When the degree of mixing is good enough, the water molecules have a wide range of temperature distributions, which can create localized high temperatures before the total temperature reaches the ignition point and can trigger ignition in advance.

Water ramjet engines generally use metal-rich fuel propellants,18,19 and the practical applications of ANP additives are very wide, and the famous Russian-produced Storm (Shkval, Squall) torpedo, which can reach a water speed of 100 m/s (about 200 kts), is the most representative supervacuum torpedo using water ramjet engines.20 In order to obtain the most realistic simulation results, we adopt the working condition of this type of water ramjet engine as the simulation working condition.

2. Methods

2.1. Berendsen External Heat Bath Method

The consistency of temperature in molecular dynamics simulation is closely related to the length of the calculated temperature control,21 and in general, the shorter the length of the temperature control, the more accurate it is. At the same time, the length of temperature control is also closely related to the strength of temperature control; usually, the longer the length of temperature control, the larger the span of temperature control. The Berendsen external thermal bath method is the weakest of the temperature control methods22,23 and is more suitable for simulating simulations with a large temperature span, that is, combustion with a high degree of hot and cold mixing.

The Berendsen thermostat is a weak coupling formulationA weaker coupling formulation,24 where the system is coupled to an external heat bath with a fixed temperature in order to maintain temperature. The particle velocity is scaled at each step so that the rate of change of temperature is proportional to the difference in temperature:

2.1. 1

in which τ is the coupling parameter, which determines how tightly the bath and system are coupled together. This method gives an exponential decay of the system toward the desired temperature. The temperature change between successive time steps is given by

2.1. 2

It can be seen that the rate of change of temperature is proportional to the length of time–temperature control; at this time, the proportionality factor of the speed is shown in eq 3:

2.1. 3

Due to Inline graphic using a method called the leapfrog algorithm to integrate the time, in calculation, τ is used to adjust the strength of the coupling as an empirical parameter, so its value must be chosen carefully. As τ → ∞, the Berendsen thermostat becomes stable. In the calculation process, only one microcanonical ensemble is sampled, and the temperature fluctuation will increase continuously until the appropriate value of the microcanonical ensemble is reached. Of course, the proper value of a microcanonical ensemble can never be reached. The opposite of that is that too small of a τa value will lead to unrealistic low-temperature fluctuations. If τ is chosen to be the same as the time step δt, then the Berendsen thermostat is simply the speed scaling.

2.2. Relevant Parameters of Engine Operating Conditions and Simulated Environment

As shown in Table 1, the main parameters of the simulation conditions are predicted and their origins. Among the four parameters, the pressure value has a decisive influence on the construction of the model, which is the most important and needs to be determined before modeling.

Table 1. Main Parameters of Simulated Working Conditions.

parameter pressure adiabatic temperature ignition temperature melting temperature
value 2.5 MPa 3800 K <1000 K >600 K
references (25) (26,27) (28) (29)
Open in a new tab

Referring to the performance data of the “Blizzard” super cavitating torpedo, the total pressure formula used at a speed of 100 m/s in 10 m water depth is calculated as eq 4:

2.2. 4

The total pressure at the inlet can be obtained as 5.2509 MPa, and the combustion chamber pressure is taken as 2.5 MPa.25 Aluminum and water vapor can reach an adiabatic temperature of about 3700 K at a high pressure of 2.5 MPa26 (initial temperature unknown in the original Figure 1). The adiabatic temperature at an initial temperature of 698 K is 3100 K at an equivalence ratio of 127 (original Figure 6 pressure unknown). The adiabatic temperature for the combined reference combustion should be 3800 K, and both acid pyrolysis and combustion temperatures can be referenced in this temperature range.

Figure 1.

Figure 1

Open in a new tab

Initial model of the ReaxFF force-field verification simulation.

Figure 6.

Figure 6

Open in a new tab

Comparison of adsorption numbers of different coating degrees.

The ignition temperature is much earlier, and the ignition point of aluminum nanopowder decreases with diameter, and the ignition temperature is already as low as about 1000 K when the particle size decreases to 10 nm28 (Figure 2 in the original paper). Aluminum/water vapor combustion above and below 750 K is more representative,30,31 and we believe that the temperature at which premixing of the ANP with the water vapor occurs should be in the range between the boiling point of water and aluminum, that is, 400/750 K.

Figure 2.

Figure 2

Open in a new tab

Change curve of the carbon molecular chain.

The melting point of aluminum nanopowder also decreases with the decrease in diameter, Liu et al. from the Harbin Engineering University studied the melting point of aluminum nanoparticles below 10 nm,29 which is limited by the performance of the computer only to the upper limit of the calculation up to 4 nm, and the molecular dynamics simulation shows that the melting point of 4 nm aluminum nanopowder is around 600 K, based on the trend of the melting point of 6 nm nanopowder around 700 K. The melting point of 6 nm nanopowder is around 700 K, and the melting point of 6 nm nanopowder is around 700 K.

3. Molecular Dynamics Simulation

3.1. Molecular Pyrolysis Simulation of Palmitic Acid

Before carrying out the combustion simulation of chondritic acid-coated nano aluminum powder, it is necessary to carry out the force-field applicability validation simulation of chondritic acid with ANP. We know that chondritic acid has the chemical formula C16H32O2 or C15H31COOH containing 50 atoms with a relative molecular mass of 256.424. It is almost nonreactive with water, and it plays a complete role as a hindering agent in the combustion of the particles.

The ReaxFF force field is widely used in organic molecular simulation;32,33 in the verification of the ReaxFF force field for the pyrolysis of chondritic acid, we used a calculation domain of 100 × 100 × 100 Å3, which is calculated by the gas constant at 700 K and 2.5 MPa. There should be 235 gas molecules in the domain. Therefore, the number of stearic acids is set to be 5, and the number of water molecules is set to be 230. As shown in Figure 1a,b, the space of 80 Å above the z-axis contains 230 water molecules, and the acid molecules are 10 Å away from the water vapor region, which is flat and randomly distributed at the bottom as shown in Figure 1b.

The pyrolysis of chondritic acid in water vapor is just a simple validation simulation, which serves first to verify the validity of the force field and second to measure the pyrolysis temperature of chondritic acid, and this pyrolysis temperature is the main basis of judgment for the discussion of acid decomposition factors in combustion simulations. Since pyrolysis produces organic molecules with no uniform molecular formula at all, the pyrolysis process is discussed in the form of a carbon chain length statistic, which is the variation curve of the carbon molecular chain in Figure 2.

As shown in Figure 2, the horizontal coordinates of the curves are time versus average temperature, and the vertical coordinates are the number of carbon atoms present in molecules with different carbon chain lengths; we uniformly named the products with different chain lengths as CX, e.g., ethylene ethanol acetylene is uniformly named C2, and the pyrolysis is simulated to be heated stepwise at a temperature of 700–2700 K after a brief relaxation at a temperature of 1 × 1013 K/s with a time step of 0.1 fs and then maintained at 2700 K temperature for 300 ps.

The initial pyrolysis of the C16 molecule occurs at about 2300 K, and all decomposition is completed at 2700 K. The most abundant product is C2, followed by C3, accompanied by small amounts of C1 and C4. The reaction produces a number of intermediate products from C5 to C15, all of which are decomposed within the 2700 K environment. It can be concluded that the pyrolysis of chondritic acid occurs in the combustion temperature range, and the effect of pyrolysis of chondritic acid should be taken into account in the calculations.

3.2. Phase Transition Simulation of ANP

ANP melting is an important node in the ignition behavior, so the accuracy of the ANP phase transition simulation is highly required. For this reason, we constructed an ANP with a diameter of 6 nm used for phase transition simulations, as shown in Figure 6a, and the 6 nm ANP model contains a total of 6819 Al atoms.

Meanwhile, in order to verify the validity of the force-field file for the melting point simulation, we also added a borderless simulation with an edge length of 3 nm for the verification of validity. As shown in Figure 3b, the fcc lattice constant of Al is 4.05 Å and its three axial cycles are identical. So, we took 7 cycles to get the cube of side length 28.3465 Å which contains 1372 Al atoms.

Figure 3.

Figure 3

Open in a new tab

Phase transition simulation model.

The use of potential energy versus mean square displacement (MSD) of atoms to determine the melting point is a more general criterion for determining phase transitions, as shown in Figure 4 for 6 nm aluminum particles as well as the potential energy versus MSD curves for the simulation of borderless phase transitions of aluminum; we take the point of maximum slope of the potential energy curve as the melting point and determine the melting point of the singlet aluminum and the 6 nm aluminum particles to be 960/710 K, respectively.

Figure 4.

Figure 4

Open in a new tab

Phase transition simulation.

Comparing the boundary-free phase transition simulation with the phase transition simulation of 6 nm aluminum particles, we can find that the step in the potential energy curve of the boundary-free simulation is more pronounced, while that of the ANP is a process of a gradual increase in the slope of the potential energy curve. This is due to the presence of bare metallic bonds on the surface of pure aluminum particles, which gives the surface atoms higher energies,34 resulting in a relatively earlier melting process on the surface of the ANP, which lowers the melting point of the ANP as a whole.

3.3. Water Molecule Adsorption Simulation of Low Coating Degree

The role of the simulation of water vapor adsorption is to study the insulation efficiency of the cladding layer at low chondritic acid cladding levels using exactly the same model and working conditions except for the length of the temperature control and to use the results as a reference object for the combustion reaction so as to maximize the exclusion of irrelevant variables involved in the discussion of the combustion behavior.

For the model, we annealed the 6 nm aluminum powder of Figure 3a and used the 3410 aluminum atoms closest to the center as the aluminum core and the other 3409 aluminum atoms as the aluminum shell. As in eq 5, 10229 water molecules should be used for a water/fuel ratio of 1. After calculating the physical properties of the gases, these gases can fill a space with a side length of 351 Å by removing the volume of particles at 700 K and 2.5 MPa. Since there is no change in gas moles for the aluminum–water reaction, in this paper, we use only the NVE system for simulations.

3.3. 5

Using the volume calculation method, the full cladding state composite 6 nm chondritic acid ANP contains 886 chondritic acids, at which time each chondritic acid occupies a volume of a conical region, considering that the taper problem still requires 656-acid molecules to complete the full cladding, so we define chondritic acid cladding below 300 as low cladding.

Figure 5 shows the computational model for adsorption/combustion, where the number of stearic acid encapsulation for (a)/(b)/(c)/(d) is 0/100/200/300, respectively. As described in the previous section, the edge length is 351 Å, the diameter of the aluminum particles is 60 Å, the length of stearic acid is about 20 Å,35 the double-bonded oxygen atoms of stearic acid at the carboxyl end of the stearic acid are about 2.5 Å from the surface of the aluminum, and the distance of the closest water vapor from the center point is 60 Å, which means that the periphery of the chondritic acid still has more than 7 Å of clearance from the water vapor.

Figure 5.

Figure 5

Open in a new tab

Adsorption/combustion models with different coating degrees.

In the low-coverage simulations we have completed, ignition behavior occurred with aluminum nanopowder coated with 100 molecules of chondritic acid versus 300 molecules of chondritic acid. We did not experience any more ignition behavior after the temperature control duration was shortened to 0.1 fs for these two calculations, which reflects the effectiveness of using the temperature control duration to increase the temperature span of water molecules. The physical adsorption of water onto the aluminum surface is mainly influenced by van der Waals and Coulomb forces, while the temperature span of the water molecules has a negligible effect on the adsorption; therefore, our simulations of the adsorption efficiencies of aluminum nanopowders with different degrees of cladding provide comparisons for combustion under equivalent conditions.

We used whether the O atom in the middle of the water molecule comes within 2 Å of the Al atom as a criterion for whether the adsorption is complete, as shown in Figure 6, and we found that of the three arithmetic cases in the presence of chondritic acid encapsulation, there is a general tendency for the more adequately encapsulated aluminum powder particles to have lower adsorption efficiencies. More unusual is the uncoated count, where the efficiency of the uncoated aluminum powder adsorption is severely lagged at the beginning of the adsorption period, although the number of adsorbed aluminum powders without coating ultimately exceeds that of the three acid-containing coated counts. We attribute this efficiency lag to the fact that van der Waals forces are the main force for adsorption when water molecules are far away from the aluminum powder.35 While chondritic acid increases the van der Waals force, it also further shortens the distance to the water molecules, which is why the phenomenon of the lagging adsorption efficiency of uncoated aluminum powder is formed.

We compare the trends in the number of Al atoms contained in the metallic Fcc lattice at four levels of cladding over a 300 ps time period. The stable oxide film on the aluminum surface is about 28 Å in the natural environment,36 and the oxidation reaction proceeds spontaneously upon exposure to water molecules. As shown in Figure 7, the number of Fcc structure atoms at the time of origin is an isotropic series due to the disruption of the metal structure by adsorption, with an average decrease of approximately 300 Fcc structure atoms for every 100-acid molecules, consistent with the conclusion that adsorption of one soft acid affects three Al atoms.35

Figure 7.

Figure 7

Open in a new tab

Comparison of the Fcc lattice number with different coating degrees.

The oxidation trend in Figure 7 is roughly consistent with the adsorption trend in Figure 6. Uncoated ANP oxidizes with a relative lag but ends up oxidizing to the highest degree. 300-acid molecule coating is the opposite. It is noteworthy that the efficiencies of 100- and 200-acid molecular coatings are roughly parallel, with a consistent gap in progress. This reflects the lag of the oxidation reaction relative to the adsorption behavior.

Combining Figures 6 and 7, we find that the low chondral acid coating has a certain effect on the isolation of water molecules from adsorption and oxidation, but the effectiveness is not high, and the simulation results are closer to those of the wet chemical method,37,38 which is not effective because the coating layer contains a certain amount of solution content during the coating process, which disappears during the drying process.39

Energy conversion in the combustion process is the process of converting atomic potential energy to kinetic energy, as shown in Figure 8a. Due to the difference in potential energy caused by different atomic weights, the potential energy change curves of different coating degrees show a parallel relationship on the whole, and the overall decline trend of potential energy in the adsorption process is similar.

Figure 8.

Figure 8

Open in a new tab

Adsorption energy curves of different coating degrees.

As can be seen from Figure 8b, the potential energy gap between different coating degrees is mainly caused by the van der Waals energy. Since van der Waals energy is greatly affected by the atomic weight, and the acid number of the palmitic acid layer is the main difference in modeling, such a gap cannot be avoided. We can see in Table 2 that there is an anomaly in the noncoated example of van der Waals energy variation, and the van der Waals energy variation is positive, which is obviously unreasonable. Combined with the simulation experience, we believe that this is caused by too compact of a particle model and insufficient initialization. Overall, the change in the van der Waals energy is small and almost negligible compared to the change in the Coulomb energy.

Table 2. Energy of Different Coating Degrees.

  coating number 0 100 200 300
energy variation (kcal/mol) potential energy –236,730 –248,970 –258,770 –245,660
VDW energy 14,930 –6840 –14,800 –33,600
coulomb energy –247,811 –242,123 –243,975 –212,058
Open in a new tab

The decisive factor in the adsorption simulation is the Coulomb force. As shown in Figure 8c, the reaction process of 100-acid coating and 200-acid coating is similar, while the reaction of 300-acid coating will have obvious obstacles. So, this can mean that there is a threshold; even if the acid coating reaches a certain proportion, the coating effect will change qualitatively.

As shown in Table 2, the change in potential energy does not exactly coincide with the trend of adsorption and oxidation; for example, the uncoated arithmetic case has the highest degree of adsorption and oxidation and the smallest change in potential energy instead. In order to gain a deeper understanding of the potential energy change factors, we discussed the trend of van der Waals energy and Coulomb energy and added the amount of change of van der Waals energy and Coulomb energy to Table 1 as well, which also confirms the conclusion that the Coulomb force is the main factor of adsorption.

3.4. Premixed Combustion Simulation of Low Coating Degree

The premixed ignition behavior is shown in Figure 9, with local area temperatures exceeding 2000 K. Since the cladding layer hinders water molecule adsorption to a certain extent, ignition requires a certain amount of water molecules to be adsorbed on the aluminum surface. Once the combustion depletes the adsorbed water molecules, the combustion intensity decreases, reflecting the chain reaction.

Figure 9.

Figure 9

Open in a new tab

Combustion diagram of 100 palmitic acid combined with ANP at 50 ps/450 K.

As can be seen in Figure 9, the atomic temperature range of water molecules is large, a certain proportion of the atomic temperature of water molecules can be maintained above 1000 K, and a few atoms are even close to 2000 K, which is the key factor to induce premix ignition behavior.

The premixed ignition behavior is achieved by decreasing the frequency of temperature control; the arithmetic example containing 100/300 stearic acid cladding underwent premixed ignition, using the software OVITO, we can see that the 100/300-acid cladding combustion ignition occurs at 35/39.5 ps, that is, a total temperature of 435/439.5 K, which very accurately responded to the diffusivity curve in Figure 10.

Figure 10.

Figure 10

Open in a new tab

Diffusivity of ANP combustion.

We compared the diffusivity of the aluminum core with that of the aluminum shell, as shown in Figure 10a for the ANP core. The nuclear diffusivity is almost zero when no combustion occurs; it rises linearly when combustion behavior occurs, and the magnitude of the diffusivity is determined by the intensity of the combustion. We determine that localized melting occurs in the shell layer and diffuses first in the nucleus due to the metal’s thermal conductivity advantage. Figure 10b shows the diffusivity of the ANP shell layer before ignition occurs; the shell layer receives water molecule adsorption influencing the diffusivity to maintain a certain rate, after ignition occurs, the aluminum core diffusivity will very quickly exceed the aluminum shell, 175 ps after the shell layer diffusivity also rises linearly, the 100-acid cladding degree shell layer near 280 ps exceeds the aluminum core, this trend is the cladding layer off with the reaction of the surface layer caused by the reaction.

As can be seen from Figure 10, the intensity of combustion of the 300-acid coating is much higher than that of the 100-acid coating, which is only the manifestation of randomness.

As shown in Figure 11, we can find that the destruction of the aluminum metal Fcc lattice by combustion is much more efficient than adsorption when the same degree of cladding. In this simulation, the 300-acid molecule-coated case has a more intense combustion reaction, so the 300-acid-coated combustion case has the most intense destruction of the Fcc structure and disappears around 250 ps, becoming the only case that destroys the metal lattice. The oxidation efficiency of the 100-acid-coated case in the adsorption case eventually exceeds that of the 300-acid-coated case, and the combustion case is essentially the same as the 100-acid-coated case in terms of metal structure destruction due to the intense combustion of the 300-acid-coated case.

Figure 11.

Figure 11

Open in a new tab

Damage of aluminum powder particle structure by combustion.

The change in potential energy most accurately reflects the energy produced by combustion, as shown in Figure 12, the potential energy change of the combustion algorithm is obviously more rapid than the adsorption, in which the 300-acid combustion ends up being 149,847.3 kcal/mol lower than the adsorption potential energy, relative to the 100-acid combustion/covering potential difference of 40,895.9 kcal/mol, and the 300-acid combustion is about 100-acid combustion intensity The intensity of 300-acid combustion is about 3.66 times that of 100-acid combustion.

Figure 12.

Figure 12

Open in a new tab

Potential energy curve of the combustion system.

We have done a lot of research on the radial distribution of various variables, including density, temperature, internal stress, charge, kinetic energy density, potential energy density, etc.; however, there are few results worthy of in-depth research and analysis. As shown in Figure 13 for the radial distribution of temperature at each time point, our research method is to take the center point of the particles as the origin and divide a circular area with a radius of 110 Å into 55 equal parts to form a radial layer distribution with a thickness of 2 Å. The radial distribution of the temperature at each time point is shown in Figure 13 for the radial distribution of the temperature at each time point.

Figure 13.

Figure 13

Open in a new tab

Radial distribution of temperature at each time point.

The data within 5 Å as shown in Figure 13 are extremely jumpy, which is because the closer to the origin, the smaller the atomic weight and the heavier the influence of Brownian motion on the various variables, and the physical quantities embody more randomness. Near 30 Å is the surface of the particles, and since the temperature-controlled as well as combustion regions have a low percentage of the distribution on the surface, the radial distributions of the variables do not have relatively large peaks in temperature, except for the more pronounced peaks on the surface of the 300-acid cladding in Figure 13b, where the radial distributions of the variables do not have relatively large peaks. The radial distributions of the variables are not more obvious.

We can also find that when the combustion is more moderate, such as the 100-acid-coated combustion shown in Figure 13a, the 50 ps shows an obvious high temperature on the surface of the particles near 33 Å of the curve due to the vigorous combustion at the beginning, which also affects the acid temperature distribution near 50 Å. The acid pyrolysis simulation shown in Figure 5 shows a low percentage of the acid pyrolysis distribution. The acid pyrolysis simulation shown in Figure 5 shows that this localized instantaneous high temperature can destroy the molecular structure of chondritic acid, and the obvious low temperature embodied between 33 and 50 Å reflects the gap between the acid molecule and the aluminum thermal conductivity.

The shedding of the cladding structure is very important for the combustion process, and the shedding/pyrolysis statistics of stearic acid molecules is a difficult problem. The product counting method in the pyrolysis simulation of stearic acid is extremely voluminous, the pyrolysis counting of 5 acid molecules takes several days, and the effect of more manpower input is also doubtful.

Counting the bond lengths of the oxygen and carbon atoms of the carboxy-terminal double bond has been repeatedly verified to be an effective method because past studies have shown that the double-bonded oxygen atom at the carboxy-terminal end of chondritic acid is the most important for adsorption and that the region near the double-bonded oxygen of chondritic acid forms the largest positive charge density region near the entire acid molecule, as shown in Figure 14.

Figure 14.

Figure 14

Open in a new tab

Charge density distribution of palmitic acid.

The bond length of this double bond is 1.203 Å when we optimize the conformation of Figure 14 chondritic acid in Materials Studio, and we know that the covalent radii of the carbon and oxygen atoms are 0.77 and 0.73 Å, respectively, which, together with the effect of the double bond, makes such a bond length reasonable.

However, when the oxygen of the double bond is adsorbed to the surface of the Al particles, this structure will change to some extent, and it is generally believed that the C–O double bond will weaken with the formation of the Al–O bond. However, if this weakening causes the C–O bond to break, then the entire acid molecule will lose its basis of attachment, causing soft stearic acid to fall off.

We know that the van der Waals radii of carbon and oxygen atoms are 1.72 and 1.52 Å, respectively, and if van der Waals adsorption becomes the bottom line of C–O bond adsorption, it becomes necessary to discuss a reasonable criterion of C–O bond length for judging whether adsorption exists or not.

As shown in Figure 15, we have counted the C–O bond lengths of the 100-acid and 300-acid models that have completed relaxation/adsorption/combustion, and we can clearly find that a significant portion of the C–O bonds at the completion of relaxation have bond lengths around 1.5 Å, which means that the bond properties reflect the characteristics of the covalent bonds. The 100-acid case contains 42 holding covalent bonds after relaxation, and the 300-acid case holds 98 covalent bonds after relaxation. The 100-acid example contains 42 holding covalent bonds after relaxation, and the 300-acid example holds 98 covalent bonds after relaxation.

Figure 15.

Figure 15

Open in a new tab

Bond length after relaxation/adsorption/combustion.

The other major bond length distribution does manifest itself around 3.24 Å, reflecting the characteristics of van der Waals bonds, which already outnumber covalent bonds as the dominant form of adsorption in the initial model after relaxation. After 300 ps of adsorption–combustion, the covalent bonds are reduced to a very low ratio, but this ratio is close in both the adsorption and combustion calculations. It can be seen that the intensity of the premixed combustion is less likely to destroy the strength of the covalent bonds.

The difference in the number of bonds between adsorption and combustion is even more clearly demonstrated once the bond length exceeds 3.24 Å. It is evident that combustion is very effective in destroying van der Waals adsorption. After comparison, we have taken 3.5 Å as a criterion for adsorption and used this criterion as a basis for further discussion of the efficiency of premixed combustion in destroying the structure of the cladding layer.

As shown in Figure 16, the 100-acid molecule adsorption simulation basically stabilizes the C–O bond after 100 ps, and no more shedding occurs, and the 300-acid molecule adsorption simulation of the C–O bond embodies a trend of slow increase, with a lower shedding ratio and higher adsorption effectiveness than the 100-acid molecule adsorption simulation.

Figure 16.

Figure 16

Open in a new tab

Trend of palmitic acid shedding during adsorption/combustion.

The 100-acid molecule combustion at the beginning of the shedding ratio is higher in the vicinity of 220 ps by the 300-acid molecule combustion beyond the combustion. 300-acid molecule combustion simulation combustion may be due to the initial conditions of different reasons, such as different unknown reasons, 300-acid molecule ignition is late, later ignition time makes the ignition of the time point of the water vapor adsorption more adequate, while the 300-acid molecule is to narrow the adsorption of contact surface, resulting in the ignition of the heat of the concentration concentration. This behavior accelerates the shedding and pyrolysis of the stearic acid layer, and since the combustion and the shedding process of stearic acid have a tendency to gain each other, the intensity of combustion reflects the Matthew effect.

4. Results and Discussion

Our Reaxff force field demonstrates excellent suitability in the simulation of the ANP phase change with the pyrolysis of palmitic acid, which is capable of sufficient pyrolysis in the adiabatic temperature range of aluminum to facilitate the formation of gaseous work products in the engine environment so that the palmitic acid coating cannot be a barrier to sustained combustion. Low palmitic acid coatings have limited effectiveness in isolating aluminum surfaces. If the degree of palmitic acid coating is insufficient, then the contact of water molecules with aluminum particles and the oxidation reaction of aluminum atoms cannot be avoided. The increase in the amount of acid improves the coating effect to some extent but does not achieve effective isolation.

The premixed ignition behavior is strongly influenced by the temperature distribution of the water molecules, i.e., the mixing conditions) and does not depend entirely on the overall temperature of the environment. The presence of premixed ignition creates an increased combustion flow of organically coated ANP, forming four stages of precombustion adsorption, premixed ignition, melting decoating, and full combustion. In addition, the bond length of the adsorbed carboxylate C–O double bond is one of the effective means of determining the adsorption effect, and 3.5 A can be used as an effective dividing line. Coating removal and the intensity of combustion are mutually reinforcing, so an effective coating better protects the coating structure itself.

Acknowledgments

This research was funded by Fundamental Research Funds for the Central Universities (No. 3072020CFT0203).

The authors declare no competing financial interest.

References

  1. Huang H.-T.; Zou M.-S.; Guo X.-Y.; Yang R.-J.; Li Y.-K.; Jiang E.-Z.; Li Z.-S. Study of different Al/Mg powders in hydroreactive fuel propellant used for water ramjet. Journal of Energetic Materials 2014, 32, S83–S93. 10.1080/07370652.2013.835008. [DOI] [Google Scholar]
  2. Piercey D. G.; Klapoetke T. M. Nanoscale aluminum-metal oxide (thermite) reactions for application in energetic materials, Central European. J. Energ. Mater. 2010, 7, 115–129. [Google Scholar]
  3. Meda L.; Marra G.; Galfetti L.; Severini F.; De Luca L. Nano-aluminum as energetic material for rocket propellants. Materials Science and Engineering: C 2007, 27, 1393–1396. 10.1016/j.msec.2006.09.030. [DOI] [Google Scholar]
  4. Jayaraman K.; Anand K.; Chakravarthy S.; Sarathi R.. Production and characterization of nano-aluminum and its effect in solid propellant combustion. In 45th AIAA aerospace sciences meeting and exhibit, 2007; p. 1430.
  5. Vorozhtsov A. B.; Lerner M.; Rodkevich N.; Nie H.; Abraham A.; Schoenitz M.; Dreizin E. Oxidation of nano-sized aluminum powders. Thermochimica acta 2016, 636, 48–56. 10.1016/j.tca.2016.05.003. [DOI] [Google Scholar]
  6. Saceleanu F.; Atashin S.; Wen J. Z. Investigation of the effects of phase transformations in micro and nano aluminum powders on kinetics of oxidation using thermogravimetric analysis. Phys. Chem. Chem. Phys. 2017, 19, 18996–19009. 10.1039/C7CP02692G. [DOI] [PubMed] [Google Scholar]
  7. Kwon Y.-S.; Gromov A. A.; Strokova J. I. Passivation of the surface of aluminum nanopowders by protective coatings of the different chemical origin. Appl. Surf. Sci. 2007, 253, 5558–5564. 10.1016/j.apsusc.2006.12.124. [DOI] [Google Scholar]
  8. Kwon Y.-S.; Gromov A. A.; Ilyin A. P.; Rim G.-H. Passivation process for superfine aluminum powders obtained by electrical explosion of wires. Appl. Surf. Sci. 2003, 211, 57–67. 10.1016/S0169-4332(03)00059-X. [DOI] [Google Scholar]
  9. Mathe V. L.; Varma V.; Raut S.; Nandi A. K.; Pant A.; Prasanth H.; Pandey R.; Bhoraskar S. V.; Das A. K. Enhanced active aluminum content and thermal behaviour of nano-aluminum particles passivated during synthesis using thermal plasma route. Appl. Surf. Sci. 2016, 368, 16–26. 10.1016/j.apsusc.2016.01.246. [DOI] [Google Scholar]
  10. Liu Y.; Sun R.; Liu P.; Xu J. Molecular Scale Simulations of Coating Palmitic Acid Molecules on Aluminum Surface. Nano 2023, 18, 2250102 10.1142/S1793292022501028. [DOI] [Google Scholar]
  11. Chu Q.; Shi B.; Liao L.; Luo K. H.; Wang N.; Huang C. Ignition and oxidation of core–shell Al/Al2O3 nanoparticles in an oxygen atmosphere: insights from molecular dynamics simulation. J. Phys. Chem. C 2018, 122, 29620–29627. 10.1021/acs.jpcc.8b09858. [DOI] [Google Scholar]
  12. Chu Q.; Shi B.; Liao L.; Zou X.; Luo K. H.; Wang N. Reaction mechanism of the aluminum nanoparticle: physicochemical reaction and heat/mass transfer. J. Phys. Chem. C 2020, 124, 3886–3894. 10.1021/acs.jpcc.9b11410. [DOI] [Google Scholar]
  13. Miller T. F.; Walter J. L.; Kiely D. H.. A next-generation AUV energy system based on aluminum-seawater combustion. In Proceedings of the 2002 Workshop on Autonomous Underwater Vehicles; IEEE, 2002; pp. 111–119.
  14. Schulte G. Fuel regression and flame stabilization studies of solid-fuel ramjets. Journal of Propulsion and Power 1986, 2, 301–304. 10.2514/3.22886. [DOI] [Google Scholar]
  15. Jiang L.; Ke Z.; Yihua X.; Haijun S.; Weigen L.. Effect of Bluff Body on Combustion Characteristics of Secondary Combustion Chamber of Solid Rocket Scramjet with Boron Contained. In Asia-Pacific International Symposium on Aerospace Technology; Springer, 2021; pp. 1199–1213. [Google Scholar]
  16. Liu P.; Liu J.; Wang M. Ignition and combustion of nano-sized aluminum particles: A reactive molecular dynamics study. Combust. Flame 2019, 201, 276–289. 10.1016/j.combustflame.2018.12.033. [DOI] [Google Scholar]
  17. Ruochen S.; Guangbao M. Influence of Alloying Elements Content on High Temperature Properties of Ti-V-Cr and Ti-Al-V Series Titanium Alloys: A JMatPro Program Calculation Study. J. Phys.: Conf. Ser. 2023, 2639, 012019 10.1088/1742-6596/2639/1/012019. [DOI] [Google Scholar]
  18. Ramakrishnan M.; Chakravarthy S. R.; Kandasamy J.; Sarathi R. Development of Nano-Al Based Highly Metalized Fuel-Rich Propellant for Water Ramjet Propulsion Applications. Propellants, Explosives, Pyrotechnics 2020, 45, 1026–1040. 10.1002/prep.201900384. [DOI] [Google Scholar]
  19. Hu J.; Han C.; Xia Z.; Huang L.; Huang X. Experimental investigation on combustion of high-metal magnesium-based hydroreactive fuels. Journal of Propulsion and Power 2013, 29, 692–698. 10.2514/1.B34675. [DOI] [Google Scholar]
  20. Ashley S. Warp drive underwater. Sci. Am. 2001, 284, 70–79. 10.1038/scientificamerican0501-70. [DOI] [Google Scholar]
  21. Fu S.; Deng G.; Dong H.; Mou Y.; Hu Y.; Zhou F.; Yuan H. Numerical simulation of oil dewatering in a disc centrifuge based on PBM model. Experimental and Computational Multiphase Flow 2023, 5, 212–220. 10.1007/s42757-022-0137-7. [DOI] [Google Scholar]
  22. Hünenberger P. H.Thermostat algorithms for molecular dynamics simulations. In Advanced computer simulation: Approaches for soft matter sciences; Springer, 2005; Vol. 173, pp. 105-149. [Google Scholar]
  23. Lemak A.; Balabaev N. On the Berendsen thermostat. Mol. Simul. 1994, 13, 177–187. 10.1080/08927029408021981. [DOI] [Google Scholar]
  24. Rühle V.Berendsen and nose-hoover thermostats. Am. J. Phys. 2007, 575. [Google Scholar]
  25. He Z.; Gao Z. Q.; Gu X.; Sun X. Y.; Liu A. Dimensional design and simulation of water inlet pipeline for water ramjet. J. Solid Rocket Technol. 2022, 45, 194–199. [Google Scholar]
  26. Risha G. A.; Son S. F.; Yetter R.; Yang V.; Tappan B. Combustion of nano-aluminum and liquid water. Proceedings of the Combustion Institute 2007, 31, 2029–2036. 10.1016/j.proci.2006.08.056. [DOI] [Google Scholar]
  27. Risha G.; Huang Y.; Yetter R.; Yang V.; Son S.; Tappan B.. Combustion of aluminum particles with steam and liquid water. In 44th AIAA Aerospace Sciences Meeting and Exhibit, 2006; p. 1154.
  28. Sundaram D. S.; Yang V.; Huang Y.; Risha G. A.; Yetter R. A. Effects of particle size and pressure on combustion of nano-aluminum particles and liquid water. Combust. Flame 2013, 160, 2251–2259. 10.1016/j.combustflame.2013.04.025. [DOI] [Google Scholar]
  29. Liu J.; Wang M.; Liu P. Molecular dynamical simulations of melting Al nanoparticles using a reaxff reactive force field. Materials Research Express 2018, 5, 065011 10.1088/2053-1591/aac653. [DOI] [Google Scholar]
  30. Chen X.; Xia Z.; Huang L.; Na X.; Hu J. Experimental and Numerical Simulation of New Aluminum and Steam Vortex Combustor. Energy Fuels 2018, 32, 2458–2464. 10.1021/acs.energyfuels.7b03412. [DOI] [Google Scholar]
  31. Chen X.; Xia Z.; Huang L.; Ma L. Numerical Simulation of a Vortex Combustor Based on Aluminum and Steam. Energies 2016, 9, 1072. 10.3390/en9121072. [DOI] [Google Scholar]
  32. Sun Z.; Li K.; Bu Y.; Liang Z.; Zhan W.; Jiang C.; Zhang J. Co-pyrolysis behavior and synergistic mechanism of sub-bituminous coal and polypropylene: A ReaxFF molecular dynamics simulation. Journal of the Energy Institute 2024, 113, 101505 10.1016/j.joei.2023.101505. [DOI] [Google Scholar]
  33. Wolf M.; Vallverdu G. S. Reactive molecular dynamics simulations of plastics pyrolysis with additives: Comparison of ReaxFF branches and experimental results. J. Anal. Appl. Pyrol. 2024, 177, 106266 10.1016/j.jaap.2023.106266. [DOI] [Google Scholar]
  34. Lv F.; Qi H.; Liu P.; Liu J. Molecular dynamics simulation of the thermal pulse explosion of metal nanowire. AIP Adv. 2018, 8, 075307 10.1063/1.5037662. [DOI] [Google Scholar]
  35. Liu Y.; Sun R.; Liu P.; Xu J. Molecular Scale Simulations of coating palmitic acid molecules on aluminum surface. Nano 2022, 18, 2250102 10.1142/S1793292022501028. [DOI] [Google Scholar]
  36. Strohmeier B. R. An ESCA method for determining the oxide thickness on aluminum alloys. Surface and interface analysis 1990, 15, 51–56. 10.1002/sia.740150109. [DOI] [Google Scholar]
  37. Gromov A.; Ilyin A.; Förter-Barth U.; Teipel U. Characterization of aluminum powders: II. Aluminum nanopowders passivated by non-inert coatings, Propellants, Explosives, Pyrotechnics: An International Journal Dealing with Scientific and Technological Aspects of Energetic Materials 2006, 31, 401–409. 10.1002/prep.200600055. [DOI] [Google Scholar]
  38. Gromov A. A.; Förter-Barth U.; Teipel U. Aluminum nanopowders produced by electrical explosion of wires and passivated by non-inert coatings: Characterisation and reactivity with air and water. Powder technology 2006, 164, 111–115. 10.1016/j.powtec.2006.03.003. [DOI] [Google Scholar]
  39. Gromov A. A.; Strokova Y. I.; Teipel U. Stabilization of metal nanoparticles–a chemical approach. Chemical Engineering & Technology: Industrial Chemistry-Plant Equipment-Process Engineering-Biotechnology 2009, 32, 1049–1060. 10.1002/ceat.200900022. [DOI] [Google Scholar]

Articles from ACS Omega are provided here courtesy of American Chemical Society

RESOURCES