Investigation of Alcohol Conformer Distribution and Hydrogen Bonding in (2,2′-Difurylmethane + n-propanol or n-butanol) Binary Mixtures Using Molecular Dynamics Simulations

  • Otsile William Kgagodi
  • Foster MbaiwaEmail author
Original Article


Molecular dynamics simulations of 2,2′-difurylmethane (DFM)/n-propanol and DFM/n-butanol binary liquid mixture have been performed using the optimized potentials for liquids simulations all-atom. The density and excess molar volume were computed with DFM mole fraction ranging from 0 to 1. There is excellent agreement between the calculated and experimental density in the entire DFM composition range. Excess molar volume is negative and positive in the lower and higher mole fraction of DFM respectively, which is in accord with the experimental data (Mokate and Ddamba in J Solut Chem 35:1493–1503, 2006. The conformer distributions for n-propanol and n-butanol in neat system and in the presence of DFM were similarly studied. It was found that in both cases there is a gradual increase in the gauche conformation population as DFM mole fraction is increased from 0 to 1. Furthermore, the correlation between the dihedral angles of n-propanol and n-butanol with the oxygen of DFM-acidic hydrogen (ODFM–Halc) radial distributions are investigated. The correlation is positive for trans conformation and negative for gauche conformation, which is attributed to the structural ease of hydrogen bond donation by both alcohols. Finally, the spatial distribution of DFM and n-butanol around a DFM molecule are examined. It is revealed that both molecules orient uniquely around the DFM molecule, which modifies the intermolecular interactions.


Molecular dynamics Hydrogen bonding Conformer Excess molar volume Radial distribution functions 

1 Introduction

2,2′-difurylmethane (DFM) is an organic molecule consisting of two furan rings separated by a methylene bridge via one of the alpha carbons. It is a colorless liquid at room temperature. DFM is widely used as a flavorant and odorant in the food industry [2]. Many industrial chemicals contain small amounts of DFM as a side product. For example, polyfurfuryl alcohol polymer (PFA) resin has been found to contain many chemicals one of which is DFM [3]. A review of the uses and applications of furylarylmethanes in general can be found in Ref. [4]. Among those, include applications as monomers and linkers in the synthesis of industrially important polymers and resins. DFM and its derivatives have been used in synthesis of calixarenes which are useful as sensors for ion selective chromatography [2]. Furylarylmethanes are favored as starting raw materials for the production of polymers because they can be made from furfuryl alcohol which can be obtained from agricultural waste [4]. DFM can also be made furfuryl alcohol thus it can be considered a greener solvent. Since DFM is a polar aprotic solvent, it can be used to modify the polarity of alcohols. Motivated by this, Mokate and Ddamba probed the volumetric properties of binary mixtures of DFM with methanol in a temperature range of 288.15–308.15 K [5]. They later studied the volumetric properties of DFM with aliphatic alcohols up to n-hexanol under the same conditions [1, 6]. In terms of excess molar volumes at 298 K, lower chain alcohols (methanol and ethanol) show volume contraction at almost all compositions, while long chain alcohols (pentanol and hexanol) largely show volume expansion at all compositions. Propanol and butanol show both volume contraction (at low DFM composition) and volume expansion (at high DFM composition) [1]. Few binary mixtures exhibit both volume contraction and expansion [7, 8]. The existence of such behavior shows the presence of both negative and positive volume contributing factors and dominance of one over the other depending on the mixing ratios. Negative contributing factors include dipole–dipole interactions between the different molecules and geometrical interstitial accommodation into each other’s cavities whereas positive contributions include repulsions within the system and unfavorable molecular packing [1]. In the case of DFM/n-alcohol mixtures, it has also been suggested that the observed changes in excess molar volumes arise due to changes in the proportion of different conformers of the alcohol molecules as the mole fraction of DFM increases [1]. This can easily be investigated using molecular dynamics.

In our previous paper we used molecular dynamics and ab initio studies to investigate details of the DFM/n-propanol binary mixture [9]. Our results showed evidence of hydrogen bonding between the acidic proton in n-propanol and the DFM oxygen atoms. The effects of mixing DFM and n-propanol on each of the two molecules were investigated mainly using radial distributions functions. The results showed that no significant changes in the local structure of both molecules occur upon mixing at different compositions, indicating that the solvents do not interact strongly. However we did not study the distribution of n-propanol conformers in that system. In this paper, the DFM/n-butanol binary mixture is studied. Due to an extra CH2 group, n-butanol has more conformational isomers than n-propanol. For both alcohols, the effect of DFM concentration on the distribution of conformers is investigated.

2 Methods

For both DFM/n-propanol and DFM/n-butanol binary liquid mixtures, molecular dynamics simulations were performed using the DLPOLY 4.08 package [10]. The optimized potentials for liquid simulation All Atoms (OPLS-AA) force field was used to describe the systems [11, 12]. Initial configurations were generated using Packmol [13]. Analysis of the trajectories was performed using TRAVIS [14]. The dynamics were performed in the Isothermal isobaric ensemble at 298.15 K and 1.0 atm using the Nose–Hoover thermostat and barostat, whose relaxation times were set to 0.1 ps and 0.5 ps respectively. The leapfrog-verlet algorithm was used for integration of equations of motion with a time step of 1.0 fs. The cut-off distance was set to 12 Å. The smooth particle mesh ewald summation was used for treating long range interactions. For each mixture, a total of 1000 molecules were placed in a cubic box. The system was then equilibrated for 1 ns and production dynamics were collected over 3 ns using a time step of 1.0 fs. This time was considered enough to capture the full dynamics given that a 15 ns long run (5 ns equilibration and 10 ns production run) using 900 DFM/100 n-butanol mixture gave essentially the same results.

3 Results and Discussions

The densities of the mixtures were calculated from the average volumes of the cubic boxes of the periodic boundaries and the masses of the mixtures. Plots of density of the DFM/n-butanol mixtures at various composition are shown in Fig. 1. Densities of DFM/n-propanol mixtures have been presented before [9]. There is good agreement of the molecular dynamics calculated density with the experimentally obtained results. From determined densities, excess molar volumes can be calculated as follows:
$$ V_{m}^{e} = \frac{M}{\rho } - \sum {\left( {\frac{{x_{i} M_{i} }}{{\rho_{i}^{*} }}} \right)} $$
where M is the molar mass of the mixture, ρ is the density of the mixture, and \( x_{i} \), \( M_{i} \) and \( \rho_{i}^{*} \) are the mole fraction, the molar mass as well as the density of the pure liquid component i respectively. In Fig. 2, the experimental and calculated excess molar volumes are also shown. Given the sensitivity of excess molar volume and the small deviations from ideal behavior, the discrepancy between the two data sets is expected. This inability of OPLS-AA to accurately reproduce the excess molar volume was noted in our previous study [9]. Perera et al. [15] and Cui et al. [16] have also reported some discrepancies with excess molar energies and excess molar volumes in their systems respectively using this force field. The calculated results however still capture the overall trend of the experimental data. At lower mole fractions of DFM, the volume contracts while at high DFM mole fractions there is volume expansion.
Fig. 1

Simulated (red diamonds) and experimental (black circles) [1] densities of DFM/n-butanol mixtures (color figure online)

Fig. 2

Simulated (red diamonds) and experimental (black circles) [1] excess molar volumes of DFM/n-butanol mixtures (color figure online)

3.1 Conformer Distribution in Pure Alcohols and in Mixtures

n-propanol has five possible conformers due to the two torsional angles, C–C–C–O and C–C–O–H. On the other hand, n-butanol has one more torsional angle, C–C–C–C, which leads to more unique conformers. Thus, it has a total of fourteen conformers. The common practice for naming the conformers is to use two letters (for n-propanol) or three letters (for n-butanol) notation to denote the two or three dihedral angles which can be either gauche (G or G′ g or g′) or trans (T or t). Lower case letters represent the C–C–O–H torsion and the primed letters are for negative dihedral angles. For example, the Tg’ conformer of n-propanol has dihedral angles of 180° for the C–C–C–O torsion, − 60° (or 300°) for C–C–O–H torsion. The GGt conformer of n-butanol has dihedral angles of 60°, 60° and 180° for the C–C–C–C, C–C–C–O and C–C–O–H torsions respectively. Visuals of the five n-propanol conformers and fourteen n-butanol conformers can be found in Ref. [17] supporting information. The relative amounts of the conformers in equilibrium systems were calculated by integrating the appropriate dihedral distribution functions. The integrations were from 0° to ~ 120° for the G/g configurations, ~ 120° to ~ 240° for the T/t configurations and ~ 240° to 360° for the g’ configurations. From the C–C–C-O dihedral angle distributions of pure n-propanol the trans configurations have a slight dominance (34.5%) over the gauche configurations which together contribute 65.5%. In the C–C–O–H dihedral angle g and g’ configurations each contribute 31% for a total of 62% and the trans configurations contribute the remaining 38%. This is comparable to recent Raman spectroscopy studies by Chen and co-workers which showed that the gauche configurations in C–C–O–H contribute 67% [17]. The same authors also obtained 63% using molecular dynamics calculations. Thus, in totality the most predominant conformer is Tt. Recent high level ab initio calculations suggest the Gt conformer is the global minimum followed by the Tt conformer lying 0.07 kcal/mol above it [18].

In the case of n-butanol gas phase ab initio studies using HF/6-31G [19] and M08-HX/6-31 + G(2df, 2p) [20] place the TTt conformer 0.17 kcal/mol and 0.32 kcal/mol respectively above the TGt which is the most stable. The latter predicts two other conformers of lower energy lie below the TTt. However, from our calculations the TTt conformer dominates at all concentrations of DFM.

The distribution of the conformers was similarly studied in the presence of 2,2′-difurylmethane. Figure 3 shows the percentage amount of the gauche conformers in n-propanol and n-butanol containing various amounts of 2,2′-difurylmethane. In both cases the percentage amount of the gauche conformers increases almost gradually. The small increase in the relative population of the gauche conformers indicates that they are easier to incorporate into the DFM network especially at higher DFM concentrations. Relative to the neat n-propanol, at 90% DFM, the difference in composition is about 4% for the CCOH and 6% for the CCCO configurations. The predominance of the Gt conformer persists even in mixtures until at ca 70% DFM composition wherein the predominant conformer becomes the Gg′. It must be mentioned however that at all compositions the distribution between Gg and Gg′ are nearly identical. Recent rotational spectroscopy studies have shown that in pure n-propanol at 298 K, Gg is only about 0.01 kcal/mol lower in energy than Gg’ [21].
Fig. 3

Percentage of gauche conformers in pure n-propanol and n-butanol and in mixtures of binary mixtures of DFM/alcohols

3.2 Hydrogen Bonding in Alcohols

We have previously shown that there is hydrogen bonding between oxygen atom in DFM (ODFM) and the acidic hydrogen in n-propanol (Halc) as revealed by a weak peak in the radial distribution functions between the two sites. Figure 4 shows the various Halc–ODFM RDFs DFM/n-butanol mixtures at three DFM concentrations. The percentage of hydrogen bonded DFM molecules were calculated from the radial distribution functions by integrating up to the first minima. Similarly, the percentage n-butanol molecules hydrogen bonded to n-butanol were calculated from the Halc–Oalc radial distribution functions. As shown in Fig. 5, the percentage of n-butanol molecules hydrogen bonded to other n-butanol molecules decreases with increase in the concentration of DFM. The fraction of alcohol molecules with hydrogen bonds decreases almost linearly from XDFM = 0.1 until XDFM = 0.6. Thereafter there is a sharper decrease indicating that inter-alcohol associations are getting weakened by the repulsive interactions and as well as the DFM/n-butanol hydrogen bonding. The opposite effect is observed for n-butanols hydrogen bonded to DFM molecules. At 10% DFM composition, there is virtually no n-butanol molecules which are hydrogen bonded to DFM. As the concentration of DFM increases, the number increases gradually. At 90% DFM, about 20% of the DFM molecules are hydrogen bonded to the alcohol. Strong inter-alcohol associations at low DFM concentrations are demonstrated by the high percentage of inter-alcohol hydrogen bonding. This leads to the observed low DFM/n-butanol hydrogen bonding. These observations are consistent with the observed trend in the excess molar volume of DFM/n-butanol mixtures. Hydrogen bonding has a negative contribution to the excess molar volume. Although hydrogen bonding increases between DFM and n-butanol as DFM increases, it is apparent that repulsive forces are dominant at high DFM composition. DFM/n-propanol mixtures mirror the same trend showing that in terms of hydrogen bonding to DFM, the two alcohols are not significantly different.
Fig. 4

Halc–ODFM site to site radial distribution functions for DFM/n-butanol mixtures at different DFM concentrations

Fig. 5

Simulation time averaged fraction of molecules of n-butanol molecules hydrogen bonded to n-butanol and DFM at different mixture compositions

To investigate whether hydrogen bonding plays any significant role in changing the clustering environment, correlations between alcohol dihedral angle distributions and the ODFM–Halc radial distributions functions were calculated. It was found that the correlation is positive for trans configuration and negative for the gauche configurations. This is understandable since the trans configuration is more linear allowing for a closer, less hindered approach of the alcohol molecule to oxygen atoms in DFM. An example of a correlation map is shown in the Fig. 6 for CCOH dihedral angles for DFM/n-butanol mixture at XDFM= 0.9.
Fig. 6

Correlation map for Halc–ODFM bond distance and n-butanol C–C–O–H dihedral angle for XDFM = 0.9

3.3 Analysis of the First Solvation Shell

For strongly associating liquids, the first coordination shell is highly incompressible, with their structure remaining relatively unchanged across various densities [22]. Since the nature of the coordination shells of strongly associating liquids is a direct consequence of their propensity for hydrogen bonding, we have examined the first solvation shell. Coordination numbers of the first solvation shell of the mixtures were calculated from the radial distribution functions of center of masses for the molecules of interest according to Eq. (2).
$$ N_{ij} = 4\pi \rho_{j} \mathop \int \limits_{0}^{{R_{min} }} g_{ij} \left( r \right)r^{2} dr $$
\( N_{ij} \) represents the number of species j surrounding species i in a shell extending from zero to \( R_{min} \) (the first minimum in the RDF) and \( \rho_{j} \) is the number density of species j. \( g_{ij} \left( r \right) \) is the center of mass radial distribution between species i and j. For a binary mixture, coordination numbers, N11, N12, N21 and N22 can be calculated, where the first numeric refers to the reference molecule and the second numeric is for the observed (or surrounding) species. For the DFM (1) and n-butanol (2) mixtures the variation of the coordination numbers with increasing DFM concentration is shown in Fig. 7. The coordination numbers increase with an increase in the concentration of the solvating species. The excess mole fractions \( X_{DFM}^{E} \), which is the difference between bulk mole fraction and the mole fraction in the first solvation shell (local mole fraction) was calculated using the formula
$$ X_{DFM}^{E} = X_{DFM}^{bulk} - \left( {\frac{{N_{11} + N_{21} }}{{N_{11} + N_{21} + N_{12} + N_{22} }}} \right) $$
A plot of the excess mole fractions at different DFM compositions is shown in Fig. 8. As in the case of DFM/n-propanol mixtures, \( X_{DFM}^{E} \) remarkably mirrors the excess molar volumes indicating that first solvation shell has a huge impact of the observed bulk properties. For the mole fraction of DFM ranging from 0.1 to 0.6, the \( X_{DFM}^{E} \) assume large negative values. This indicates stronger mutual interactions than self-interactions of DFM and n-butanol especially at 0.3 DFM mole fraction. From 0.7 to 0.9 DFM mole fractions, the \( X_{DFM}^{E} \) values are either positive or very close to 0. In this case, self-interactions are favored than mutual interactions.
Fig. 7

Variation of coordination numbers with concentration of DFM in n-butanol

Fig. 8

Excess mole fraction of DFM in the first coordination shell as calculated according to Eq. (3)

Figure 9 shows the spatial distribution plots of DFM and n-butanol around DFM reference molecule. It is clear that alcohol molecules (represented by blue surface) are mostly located above and below the DFM furan ring. On the other hand DFM molecules are mostly located on the sides of the DFM reference molecule’s furan rings. This arrangement likely strengthens the interaction between the DFM and alcohol molecules and reduces the repulsion between DFM–DFM furan rings.
Fig. 9

Spatial distribution of molecules around one DFM molecule XDFM = 0.5. The n-butanol density is represented by blue and DFM by red cloud. In both cases the isovalue was set to 90% of the maximum density

4 Conclusions

Molecular dynamics simulations using OPLS-AA force field was used to investigate the thermodynamics and structural properties of DFM/n-butanol binary mixture. The calculated results satisfactorily reproduce the experimental density. In case of the excess molar volume there is a reasonable discrepancy with the experimental data. Nevertheless, the calculated excess molar volume trend is consistent with experimentally calculated results. Excess molar volumes exhibit negative deviation for lower DFM mole fractions and positive deviation for higher mole fractions.

For structural investigations, we have calculated the conformer distributions, hydrogen bond interactions and first solvation shell populations for DFM/n-propanol and DFM/n-butanol mixtures. The results indicate that there is a small increase in gauche conformation for both alcohols as the DFM mole fraction is increased from 0 to 1. This suggests that gauche conformers are more able to incorporate into the DFM local network at higher DFM mole fractions. Correlations between alcohol conformer and the alcohol/DFM hydrogen bonding show that there is positive correlation for trans conformations and negative values for gauche conformations. This is not surprising since trans conformers have structurally less hindered with respect to approaching the DFM oxygen atoms for hydrogen bonding. The excess mole fractions of DFM/n-propanol indicate that the first coordination shell has a huge impact on the overall observed bulk properties. Finally, the spatial distribution of DFM and n-butanol around DFM molecule shows n-butanol molecules are mostly located above and below DFM furan ring, while DFM molecules are mostly on the sides. This arrangement maximizes the interaction energy between DFM and n-butanol and minimizes repulsion between DFM–DFM furan ring.



The Authors would like to thank the Botswana International university of Science and Technology for supporting the research and scholarship for OWK. We also would like to thank University of Botswana CSHPC for use of their computational resources.

Compliance with Ethical Standards

Conflict of interest

On behalf of all authors, the corresponding author states that there is no conflict of interest.


  1. 1.
    Mokate O, Ddamba WAA (2006) Volumetric properties of (difurylmethane + alkan-1-ol) binary mixtures at 298.15 K. J Solut Chem 35:1493–1503. CrossRefGoogle Scholar
  2. 2.
    Khusnutdinov RI, Baiguzina AR, Smirnov AA, Mukminov RR, Dzhemilev UM (2007) Furfuryl alcohol in synthesis of levulinic acid esters and difurylmethane with Fe and Rh complexes. Russ J Appl Chem 80:1687–1690. CrossRefGoogle Scholar
  3. 3.
    Iroegbu AO, Hlangothi SP (2018) Effects of the type of catalyst on the polymerisation mechanism of furfuryl alcohol and its resultant properties. Chem Afr. Google Scholar
  4. 4.
    Butin AV, Stroganova TA, Kul’nevich VG (1999) Furyl(aryl)methanes and their analogs (review). Chem Heterocycl Compd 35:757–787. CrossRefGoogle Scholar
  5. 5.
    Mokate O, Ddamba WAA (2005) Volumetric properties of difurylmethane in methanol from 288.15 to 308.15 K. J Solut Chem 34:1327–1339. CrossRefGoogle Scholar
  6. 6.
    Mokate O, Ddamba WAA (2008) Effect of temperature on the volumetric properties of [difurylmethane + (C2–C6) Alkan-1-ol] Binary systems: analyses of new and literature density data in the temperature range 288.15–308.15 K. J Solut Chem 37:331–350. CrossRefGoogle Scholar
  7. 7.
    Grgurić IR, Tasić AŽ, Djordjević BD, Kijevčanin ML, Šerbanović SP (2002) Excess molar volume of the acetonitrile + alcohol systems at 298.15 K. Part I: density measurements for acetonitrile + methanol, + ethanol systems. J Serb Chem Soc 67:581–586. CrossRefGoogle Scholar
  8. 8.
    Grgurić IR, Kijevčanin ML, Djordjević BD, Tasić AŽ, Šerbanović SP (2003) Excess molar volume of acetonitrile + alcohol systems at 298.15 K. Part II: correlation by cubic equation of state. J Serb Chem Soc 68:47–56. CrossRefGoogle Scholar
  9. 9.
    Kgagodi OW, Mbaiwa F (2017) Molecular dynamics study of 2,2′-difurylmethane and n-propanol binary mixture. J Mol Liq 227:366–372. CrossRefGoogle Scholar
  10. 10.
    Todorov IT, Smith W, Trachenko K, Dove MT (2006) DL_POLY_3: new dimensions in molecular dynamics simulations via massive parallelism. J Mater Chem 16:1911–1918. CrossRefGoogle Scholar
  11. 11.
    McDonald NA, Jorgensen WL (1998) Development of an all-atom force field for heterocycles. Properties of liquid pyrrole, furan, diazoles, and oxazoles. J Phys Chem B 102:8049–8059. CrossRefGoogle Scholar
  12. 12.
    Jorgensen WL, Maxwell DS, Tirado-Rives J (1996) Development and testing of the OLPS all-atom force field on conformational energetics and properties of organic liquids. J Am Chem Soc 118:11225–11236. CrossRefGoogle Scholar
  13. 13.
    Martínez L, Andrade R, Birgin EG, Martínez JM (2009) PACKMOL: a package for building initial configurations for molecular dynamics simulations. J Comput Chem 30:2157–2164. CrossRefGoogle Scholar
  14. 14.
    Brehm M, Kirchner B (2011) TRAVIS: a free analyzer and visualizer for Monte Carlo and molecular dynamics trajectories. J Chem Inf Model 51:2007–2023. CrossRefGoogle Scholar
  15. 15.
    Perera A, Zoranić L, Sokolić F, Mazighi R (2011) A comparative molecular dynamics study of water–methanol and acetone–methanol mixtures. J Mol Liq 15:52–59. CrossRefGoogle Scholar
  16. 16.
    Cui S, de Almeida VF, Khomami B (2014) Molecular Dynamics Simulations of Tri-n-butyl-phosphate/n-dodecane mixture: thermophysical properties and molecular structure. J Phys Chem B 118:10750–10760. CrossRefGoogle Scholar
  17. 17.
    Chen L, Zhu W, Lin K, Hu N, Yu Y, Zhou X, Yuan LF, Hu SM, Luo Y (2015) Identification of alcohol conformers by Raman spectra in the C–H stretching region. J Phys Chem A 119:3209–3217. CrossRefGoogle Scholar
  18. 18.
    Kahn K, Bruice TC (2005) Focal-point conformational analysis of ethanol, propanol, and isopropanol. ChemPhysChem 6:487–495. CrossRefGoogle Scholar
  19. 19.
    Ohno K, Yoshida H, Watanabe H, Fujita T, Matsuura H (1994) Conformational study of 1-butanol by the combined use of vibrational spectroscopy and ab-initio molecular-orbital calculations. J Phys Chem 98:6924–6930. CrossRefGoogle Scholar
  20. 20.
    Seal P, Papajak E, Yu T, Truhlar DG (2012) Statistical thermodynamics of 1-butanol, 2-methyl-1-propanol, and butanal. J Chem Phys 136:034306. CrossRefGoogle Scholar
  21. 21.
    Kisiel Z, Dorosh O, Maeda A, Medvedev IR, De LF, Herbst E, Drouin BJ, Pearson JC, Shipman ST (2010) Determination of precise relative energies of conformers of n-propanol by rotational spectroscopy. Phys Chem Chem Phys 12:8329–8339. CrossRefGoogle Scholar
  22. 22.
    Cipcigan F, Sokhan V, Martyna G, Crain J (2018) Structure and hydrogen bonding at the limits of liquid water stability. Sci Rep 8:1718. CrossRefGoogle Scholar

Copyright information

© The Tunisian Chemical Society and Springer Nature Switzerland AG 2019

Authors and Affiliations

  1. 1.Department of ChemistryUniversity of BotswanaGaboroneBotswana
  2. 2.Department of Chemical and Forensic SciencesBotswana International University of Science and TechnologyPalapyeBotswana

Personalised recommendations