Structure, thermal stability, antioxidant activity and DFT studies of trisphenols and related phenols

Article abstract of DOI:10.1016/j.ica.2017.07.022

Two kinds of trisphenols have been successfully synthesized and their structures were confirmed by IR spectra, ¹HNMR, ¹³CNMR, mass spectra and X-ray diffraction. They exhibited better thermal stability than both monophenol and bisphenols due to their higher molecular weight. Moreover, their antioxidant activities have been investigated in lubricant oil using PDSC and RBOT. The results showed that the o-trisphenol 3b exhibited the best antioxidant activity while the p-trisphenol 3a was the worst. In addition, their relationship between structures and properties has been further explored by a series of DFT calculations including the BDE values, the IP values and the Gibbs free energy barriers for the reaction between phenols and methylperoxyl radicals.

Full text of DOI:10.1016/j.ica.2017.07.022

Accepted Manuscript
Research paper
Structure, thermal stability, antioxidant activity and DFT studies of trisphenols
and related phenols
Shasha Yu, Yishan Wang, Yujie Ma, Limin Wang, Jia Zhu, Shenggao Liu
PII:
S0020-1693(17)30194-9
http://dx.doi.org/10.1016/j.ica.2017.07.022
ICA 17745
DOI:
Reference:
Inorganica Chimica Acta
To appear in:
Received Date:
Revised Date:
Accepted Date:
16 February 2017
19 June 2017
3 July 2017
Please cite this article as: S. Yu, Y. Wang, Y. Ma, L. Wang, J. Zhu, S. Liu, Structure, thermal stability, antioxidant
activity and DFT studies of trisphenols and related phenols, Inorganica Chimica Acta (2017), doi: http://dx.doi.org/
10.1016/j.ica.2017.07.022
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers
we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and
review of the resulting proof before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Structure, thermal stability, antioxidant
activity and DFT studies of trisphenols and
related phenols
Shasha Yu a, Yishan Wang b, Yujie Ma a, Limin Wang c, Jia Zhu b,*, Shenggao Liua,*
a Polymers and composites division, Ningbo Institute of Materials Technology and
Engineering, Chinese Academy of Sciences, Ningbo 315201, China
b College of Chemistry, Beijing Normal University, Beijing 100875, China
c School of Chemical Engineering and Environment, North University of China,
Taiyuan 030051, China
Abstract: Two kinds of trisphenols have been successfully synthesized and their
1
13
structures were confirmed by IR spectra, HNMR, CNMR, mass spectra and X-ray
diffraction. They exhibited better thermal stability than both monophenol and
bisphenols due to their higher molecular weight. Moreover, their antioxidant activities
have been investigated in lubricant oil using PDSC and RBOT. The results showed
that the o-trisphenol 3b exhibited the best antioxidant activity while the p-trisphenol
3a was the worst. In addition, their relationship between structures and properties has
been further explored by a series of DFT calculations including the BDE values, the
IP values and the Gibbs free energy barriers for the reaction between phenols and
methylperoxyl radicals.
Keywords: Antioxidant activity; Trisphenol; Thermal stability; Lubricant; DFT

1. Introduction
Phenols (especially the sterically hindered phenols) have been used as antioxidants
in a wide variety of industrial applications such as rubber processing, plastic resins,
fuel and lubricants, and others since they are always low-toxic, highly effective,
non-discoloring and eco-friendly [1-2]. A market research published by BCC
Research reported that the substituted phenols category should account for 35.6%
market share and the total market would reach to $ 1.92 billion at a five-year
compound annual growth rate (CAGR) of 5.0% by 2019 which is the fastest growing
rate of all the antioxidants [3]. 2,6-di-tert-butyl-4-methylphenol (BHT, 1) employed
from 1937 is the most frequently used antioxidant recognized as safe for application
in foods and industry [4-5]. In fact, it is unfavorable for the modern materials or
machines which need to be operated at elevated temperatures ascribed to the poor
thermal stability and antioxidant efficiency [6-7]. For example, the working
temperature of the aircraft engines is already above 200 °C while the initial
o
decomposition temperature is only 88 C [8-9]. Hence, to design and develop novel
antioxidants with superior performance at high temperatures is becoming an urgent
project.
In this regard, bisphenols with two phenolic groups in one molecule have gained a
wide range of attention in academia and industry owing to their higher antioxidant
performance and improved thermal stability. [10-15]. For example, the reactivity
toward
peroxyl
radicals
of
o-bisphenol
(6,6'-methylenebis(2-(tert-butyl)-4-methylpheno, 2b) is nearly 61-fold and 65-fold
higher than that of BHT and p-bisphenol (4,4'-methylenebis(2,6-di-tert-butylphenol,
2a) respectively on account of the less bulky substitutes and the larger stabilization of
the formed phenoxyl radical of 2b [10]. However, as for the similar compound
o-bisphenol (3,3'-di-tert-butyl-5,5'-dimethyl-[1,1'-biphenyl]-2,2'-diol, 2c) which the
two phenols are directly connected, the reactivity is practically far poorer than that of
2b and nearly identical to that of BHT since the formation of a twin intramolecular
hydrogen bond in the starting phenol largely reduces the activity of the O-H groups
[11]. It clearly demonstrates that the relative position of the two OH groups and the

linking group that connected them exert the most significant influence on the
antioxidant efficiency. Although not all the bisphenols exhibit better antioxidant
performance than BHT, their thermal stabilities have got significantly improved since
some of them such as 2a and 2b have already been applied in the industry instead.
These reports aroused our interest to investigate the properties of the trisphenols
which are composed of three phenolic moieties. First, the obviously increased
molecular weight can theoretically improve their thermal stability at elevated
temperatures [9]. Second, considering the delicate influence of the structures on the
performance of bisphenols, developing highly efficient antioxidants of trisphenols is
feasible through designing the reasonable structures. Consequently, based on the
previous reports, we have synthesized two kinds of trisphenols (3a and 3b) which the
aromatic rings are bridged by two methylene groups at different linking positions. In
this paper, other structurally related phenols including monophenol (1) and bisphenols
(2a and 2b) were also considered to further evaluate the properties of the trisphenols
(Scheme 1). Moreover, a series of DFT calculations were provided to investigate the
relationship between the structures and the properties which will be beneficial for the
future design of phenolic antioxidants.
Scheme 1. The structures of trisphenols and the related phenols.
2. Experimental
2.1. Materials

BHT and silica gel for column chromatography were purchased from Sinopharm
Chemical Reagent Co., Ltd. 4,4'-methylenebis(2,6-di-tert-butylphenol) was procured
from TCI Company. 6,6'-methylenebis(2-(tert-butyl)-4-methylphenol) was purchased
from Energy Chemical Company (Shanghai, China). 150SN was obtained from China
National Petroleum Corporation. 150N was supplied by China National Offshore Oil
Corporation. YUBASE was purchased from SK Corporation (South Korea). PAO was
provided from INEOS (England).
2.2. Characterization
1
All H and 13C NMR spectra were recorded at room temperature in CDCl3
(containing 0.03% TMS) or in DMSO-d6 on Bruker AV/ANCE-400 MHz
spectrometer. 1H NMR spectra was recorded with tetramethylsilane (δ = 0.00 ppm) or
the residual solvent peak of DMSO-d6 (δ = 2. 50 ppm); 13C NMR spectra was
recorded with CDCl3 (δ = 77.00 ppm) or the residual solvent peak of DMSO-d6 (δ =
39.52 ppm) as internal reference. High-resolution mass spectra were obtained by
using Waters Micromass GCT or Agilent Technologies 6224 TOF LC/MS mass
spectrometers. IR spectra were obtained by using a Thermo-fisher 6700 spectrometer.
Single crystal X-ray diffraction data were collected at 293(2) K for 2a and 2b on
Bruker SMART diffractometer, 130 K for 3a and 3b on Bruker APEX-II
diffractometer.
2.3. Synthesis
Synthesis Process and Structural Characterization of trisphenol 3a.
Scheme 2. Synthesis process of 3a
Trisphenol 3a was prepared with two steps by the route schematically shown in
Scheme 2.

1) Synthesis of 2,6-di-tert-butyl-4-(hydroxymethyl)phenol
Paraformaldehyde (364 mmol, 10.91 g), tBuOK (14.5 mmol, 1.63 g) and tBuOH (300
mL) were placed in a 500 mL round-bottomed flask. The flask was heated in an oil
bath at 70 °C under an inert atmosphere until the paraformaldehyde has been
completely depolymerized. After cooling to room temperature, 2,6-di-tert-butylphenol
(145.4 mmol, 30 g) was added and the obtained mixture was stirred at room
temperature for 3 h. Then 200 mL of water was added to quench the reaction. The
mixture was extracted with ethyl acetate, and the extract was washed with brine. The
organic layer was dried over anhydrous Na2SO4. The solvent was evaporated in vacuo
and the residue was purified by column chromatography on 230-400 mesh of silica
gel
(petroleum
ether/ethyl
acetate
=
10/1)
to
afford
2,6-di-tert-butyl-4-(hydroxymethyl)phenol as a white solid in 73% yield (34.4 g).
1H NMR (400 MHz, CDCl3) δ (ppm) 1.45 (singlet, 18H), 3.53 (broad, 1 H), 4.60
(singlet, 2H), 5.23 (singlet, 1H),7.19 (singlet，2H).
2) Synthesis of compound 3a
HCl (37%, 33.3 mL) was added to a solution of 4-(tert-butyl)phenol (12.6 mmol, 1.90
g) and 2,6-di-tert-butyl-4-(hydroxymethyl)phenol (25.4 mmol, 6.00 g) in DMSO (150
mL). The reaction mixture was stirred at 100 oC for 8 h in a pressure bottle. Then 200
mL of water was added to quench the reaction. The mixture was extracted with ethyl
acetate, and the extract was washed with brine. The organic layer was dried over
anhydrous Na2SO4. The solvent was evaporated in vacuo and the residue was purified
by column chromatography on 230-400 mesh of silica gel (petroleum ether/toluene =
5/1) to afford the product as a light yellow solid in 68% yield (5.08 g). 1H NMR (400
MHz, CDCl3) δ 1.27 (singlet, 9H), 1.37 (singlet, 36H), 3.88 (singlet, 4H), 4.63
(singlet, 1H), 5.05 (singlet, 2H), 7.01 (singlet, 4H), 7.04 (singlet, 2H). 13C NMR (100
MHz, CDCl3) δ 30.4, 31.7, 34.2, 34.5, 36.8, 125.3, 126.1, 127.0, 130.5, 136.2, 143.0,
150.1, 152.4. IR (cm-1): 501, 531, 611, 673, 754, 768, 794, 912, 878, 11047, 1121,
1149, 1169, 1209, 1231, 1250, 1316, 1361, 1390, 1431, 1482, 2868, 2962, 3004, 3558,
3622, 3646. HRMS (ESI) calcd for C40H62NO3 [M+NH4]+: 604.4724, found
604.4723. The structure of 3a was determined by X-ray single-crystal analysis.

Synthesis Process and Structural Characterization of trisphenol 3b.
Scheme 3. Synthesis process of 3b
Trisphenol 3b was prepared with two steps by the route schematically shown in
Scheme 3.
1) Synthesis of (5-(tert-butyl)-2-hydroxy-1,3-phenylene)dimethanol
4-(tert-butyl)phenol (20 mmol, 3.0 g) was added to an aqueous 4% NaOH solution
o
(24 mL) and the mixture stirred at 50 C to dissolve the starting material. After
cooling to room temperature, 37% HCHO solution (5.5 mL) was added and the
obtained mixture was stirred for 3 days. Then concentrated HCl (2.5 mL) was added
to quench the reaction. The mixture was extracted with chloroform, and the extract
was washed with water and brine. The organic layer was dried over anhydrous
MgSO4. The solvent was evaporated in vacuo and the residue was purified by column
chromatography on 230-400 mesh of silica gel (petroleum ether/ethyl acetate = 5/3) to
afford (5-(tert-butyl)-2-hydroxy-1,3-phenylene)dimethanol as a colorless oil in 97%
yield (4.08 g). 1H NMR (400 MHz, CDCl3) δ (ppm) 1.27 (singlet, 9H), 2.52 (broad, 2
H), 4.80 (singlet, 4H), 7.08 (singlet, 2H), 7.90 (broad, 1H).
2) Synthesis of 3b
To a solution of (5-(tert-butyl)-2-hydroxy-1,3-phenylene)dimethanol (18.5 mmol,
3.90 g) in n-heptane, 2-(tert-butyl)-4-methylphenol (130 mmol, 21.36 g) was added
and the mixture was heated until homogeneous. Then 1 mL of concentrated
hydrochloric acid was added and the mixture was heated at refluxed for 1 h using a
Dean Stark trap to remove water. After cooling to the room temperature, the mixture
was extracted with ethyl acetate, and the extract was washed with brine. The organic
layer was dried over anhydrous Na2SO4. The solvent was evaporated in vacuo and the

residue was purified by column chromatography on 230-400 mesh of silica gel
(petroleum ether/ethyl acetate = 10/1) to afford the product as a light yellow solid in
56% yield (5.21 g). 1H NMR (400 MHz, DMSO-d6) δ 1.12 (singlet, 9H), 1.34 (singlet,
18H), 2.09 (singlet, 6H), 3.85 (singlet, 4H), 6.62 (singlet, 2H), 6.81 (singlet, 2H), 6.90
(singlet, 2H), 8.11 (singlet, 2H), 8.84 (singlet, 1H). 13C NMR (100 MHz, DMSO-d6) δ
20.9, 30.0, 31.0, 31.5, 33.8, 34.8, 125.1, 125.2, 127.0, 127.8, 128.2, 128.7, 137.0,
142.2, 149.3, 150.8. IR (cm-1): 611, 638, 764, 791, 858, 878, 1131, 1153, 1206, 1241,
1288, 1315, 1375, 1453, 1484, 2868, 2957, 3316, 3372, 3583. HRMS (ESI) calcd for
C34H50NO3 [M+NH4]+: 520.3785, found 520.3781. The structure of 3b was
determined by X-ray single-crystal analysis.
2.4 Thermogravimetric analysis (TGA).
Thermogravimetric Analysis was conducted on a TGA Q500 (TA Instruments,
Newcastle, DE, U.S.A.). Samples (5 ± 0.5 mg) were heated on a platinum pan from
o
o
room temperature to 500 C at a rate of 10.0 C/min under a nitrogen atmosphere of
o
flow 60.0 mL/min for the thermal stability experiments, and 700 C at a rate of 10.0
oC/min under a compressed air of flow 60.0 mL/min for the thermo-oxidative stability
experiments. Data were analyzed using TA Universal Analysis software, version 4.5A
(TA Instruments, New Castle, DE, USA). The characteristic weight loss temperatures
comprised of the onset of weight loss temperature (TON) and the temperature at which
the maximum rate of decomposition occurred (Tmax). TON was derived from the TGA
plots and taken as the temperature at which 5% weight loss occurred (TON, 5%). Tmax
was taken as the peak maxima temperature from the corresponding DTG curves.
2.5 Isothermal PDSC.
3.0 ± 0.2 mg of oil sample was placed in a hermetically sealed aluminum pan with a
pinhole lid for interaction of the sample with the reactant gas (high-purity oxygen).
o
Oil samples were heated from 50 C to the test temperature at a heating rate of 50
oC/min before being held in an isothermal mode. After two minutes of heat
preservation, the oxygen (flow of 100 mL/min) was added in until the pressure was
3.5 MPa. When an exothermic peak of oxidation appeared, the test was finished.

Oxidation induction time (OIT) was measured from the start of the oxygen added in to
the start of the exothermic peak.
2.6 Oxidation Stability using RBOT.
The method uses 55.6 ± 0.3 g of oil with 5.0 ml of reagent water in the presence of
copper catalyst at 150 oC. In RBOT, the vessel was charged with oxygen to 620.5 kPa
pressure and rotated axially in a constant temperature at 150 °C. The pressure in the
bomb was recorded with time, and the RBOT time is the time at which the maximum
pressure of the bomb has dropped by 175.1 kPa. All oil samples were run in duplicate,
and the average RBOT times were reported.
2.7 DFT study
All calculations were performed using the Gaussian 09 program package [16]. And
the program, GaussianView [17], was used for visualization of molecular structures.
2.7.1 Calculation of BDE and IP
BDE and IP were calculated in gas-phase and n-hexane (the solvent which was
performed in the experiment). Geometrical optimizations and vibrational frequencies
consisting of molecules Ar-OH, Ar-O• and Ar-O•+ were performed using B3LYP / 6-31
G (d, p) and single-point calculations at the level of M06-2X / 6-311+G (d, p) at 25 oC
in vacuum and in n-hexane, respectively. Solvent contribution was performed
employing IEF-PCM method [18]. Besides, each optimized structures were confirmed
to be real minima by frequency analysis (Nimag = 0). Furthermore, in terms of the
computation of BDE, all possible isomeric phenolic radical forms were calculated and
the most stable isomers were selected for further consideration.
BDE can be calculated using following equation:
BDE = H (Ar-O•) + H (H•) – H (Ar-OH),
where H (Ar-O•) is the enthalpy of phenolic radical generated after H• abstraction, H
(H•) is the enthalpy of H atom, and H (Ar-OH) is the enthalpy of phenol.
IP can be calculated by the equation:
IP = H (Ar-O•+) + H (e-) – H (Ar-OH),
where H (Ar-O•+) is the enthalpy of phenolic radical cation, H (e-) is the enthalpy of

electron.
The value for gas-phase enthalpy of hydrogen atom is -0.5 hartree [19]. Gas-phase
enthalpy of electron is 3.145 kJ•mol-1 [20]. For the solvation enthalpy of the hydrogen
atom in n-hexane, where experimental value was not available, we used the average
value △solvH (H•) = 5 kJ•mol-1, because in organic solvents △solvH (H•) ≈△solvH (H2)
varies in very narrow (5±1) kJ•mol-1 range [21-23]. For the solvation enthalpy of
electron which was not available in n-hexane, it was performed at
IEF-PCM/B3LYP/6-311++G (d, p) method which proved to be a relatively robust
method when calculating solvation enthalpies of electron and proton [24-26]. And the
enthalpy changes in the process as follows:
e-gas + solventsolv  (solvent - e)-•
solv
where solventsolv is the molecule of solvent in the same solvent, and (solvent - e)-•
means that electron is attached to the molecule of solvent.
Then, the solvation enphalpy of electron is given by following equation:
△
solvH (e-) = H (solvent - e)-•solv - H (e-gas) – H (solventsolv
2.7.2 Calculation of reactions
)
The geometrical optimizations and frequency calculations were performed in
gas-phase at the B3LYP [27,28]/6-31G (d, p) [29] level at 298.15 K. The ground states
were confirmed to be real minima by frequency analysis (no imaginary frequency),
while the transition states were determined by one imaginary frequency. Besides, IRC
calculations [30, 31] were used to confirm that the transition structures properly
connect the reactants and products. To obtain more reliable relative energies,
single-point energy calculations for all optimized structures were carried at the
M06-2X [32]/6-311+G (d, p) [33] in gas-phase at 298.15 K.
3. Results and discussions
3.1 Structural analysis of trisphenols and related phenols by X-ray diffractions
and IR spectra.
Crystal structures

Fig. 1. X-ray crystal structures for bisphenols and trisphenols. Hydrogen bonds are
shown as red dashed lines.
The structures of bisphenols and trisphenols were characterized by single crystal
X-ray diffraction, shown in Fig. 1. Fig.2 provided their crystal packing diagrams. The
crystal data and refinement details are given in Table 1 and Table 2 highlights
important hydrogen bondings. Compound 2a crystallizes in the centrosymmetric
space group of the tetragonal system [34]. Two identical phenolic rings are linked by
a “CH2” group which is located para to the OH groups on the arene rings. And each
t
OH group, being nearly coplanar with the phenyl ring, is surrounded by two Bu
groups on its ortho-positions. The phenolic rings (C1/C6 and C8/C13) are twisted
with an angel of 63.78o which can be considered as a result of intramolecular steric
t
congestion between the two Bu groups. No significant intermolecular hydrogen
bonds, π-π stacking or C-H…π interactions are observed in the packing digram due to
the blocking from the bulky substituents.

Table 1. Crystal data and structure refinement for the bisphenols and trisphenols.
Identification code
Empirical formula
Formula weight
Temperature(K)
Wavelength(Å)
Crystal system
Space group
a (Å)
2a
C29H44O2
424.64
293(2)
0.71073
Tetragonal
I41/a
21.1552(12)
21.1552(12)
23.7623(17)
90
2b
C23H32O2
340.48
293(2) K
0.71073
Orthorhombic
Pna21
16.105(2)
12.7778(18)
10.1418(15)
90
3a
C40H58O3
586.86
3b
C34H46O3
502.71
130 K
130 K
0.71073
Triclinic
P-1
0.71073
Monoclinic
P121/n1
14.7284(15)
9.7166(10)
21.691(2)
90
101.619(2)
90
3040.5(5)
4
11.311(2)
12.387(3)
14.494(3)
112.184(4)
102.014(4)
94.922(4)
1808.8(7)
2
b (Å)
c (Å)
α(o)
β(o)
90
90
90
90
2087.1(5)
4
λ(o)
Volume (Å3)
Z
10634.6(14)
16
Density
(calculated)
1.061
1.084
1.078
1.098
(Mg/m3)
Absorption
coefficient (mm-1)
0.064
0.067
0.066
0.068
0.211 × 0.154 ×
0.123
0.198 × 0.156 ×
0.121
0.22 × 0.15 ×
0.08
Crystal size (mm3)
0.25 × 0.2 × 0.1
-26 ≤ h ≤ 26, -26
≤ k ≤ 23, -26 ≤ l
≤ 29
-14 ≤ h ≤16,-17 -19 ≤ h ≤ 19, -13
≤ k ≤17, -20 ≤ l ≤ k ≤ 12, -29 ≤ l
-19 ≤ h≤ 19, -15 ≤
k ≤ 15, -6 ≤ l ≤ 12
Index ranges
≤ 20
≤ 28
Reflections
collected
32314
12194
18023
26644
Independent
reflections
5243 [R(int) =
0.0556]
3072 [R(int) =
0.0353]
10838 [R(int) =
0.0612]
7882 [R(int) =
0.0452]
Final R indices
[I > 2sigma(I)]
R1 = 0.0513,
wR2 = 0.1325
R1 = 0.0815,
wR2 = 0.1510
R1 = 0.0389, wR2
= 0.1048
R1 = 0.0461, wR2
= 0.1101
R1 = 0.0673,
wR2 = 0.1329
R1 = 0.1951,
wR2 = 0.1817
R1 = 0.0490,
wR2 = 0.1148
R1 = 0.0867,
wR2 = 0.1329
R indices (all data)
Table 2. Selected hydrogen bondings for 2b and 3b (Å, o)
Compound
D-H…A
d(D-H) d(H…A)
d(D…A)
2.767(2)
<(DHA)
O(1)-H(1)...O(2)
O(1)-H(1)...O(2)
O(2)-H(2)...O(3)
0.82
0.84
0.84
1.95
1.92
1.91
172.4
171.1
170.9
2b
2.7484(15)
2.7451(16)
3b
However, compound 2b crystallizes in the centrosymmetric space group of the
orthorhombic system [34]. Two phenolic rings facing to each other with an angle of
118.06o are linked by a “CH2” group at the ortho-position of each OH group. The
t
other ortho-position and the para-position of the OH group are substituted by a Bu
group and a Me group respectively. An intramolecular hydrogen bond apparently
exists between O(1), H(1) and O(2) with an bond angle of 172.4o, resulting in the
slightly deviating of the O1-H1 bond from the plane of the corresponding phenyl ring.

In addition, no intermolecular hydrogen bonds are found in the crystal packing
diagram.
Fig. 2. The crystal packing diagrams for bisphenols and trisphenols. Intra- and inter-
molecular hydrogen bonds are shown as red and blue dashed lines respectively. The
phenyl groups that generated π-π stackings in 3b are drawn in green.
Trisphenol 3a containing three phenolic rings crystallizes in the triclinic space group
P-1. Two ortho-positions of the OH group in the central phenolic ring are both linked
with a phenolic ring through a “CH2” group. Both of the OH groups in the two side
phenolic rings are located para to the central phenolic ring and encompassed by two

tBu groups on their ortho-positions respectively, leading to the separation of the three
OH groups. An S-shaped configuration is adopted with one outer aryl group situated
above and one below the central phenolic group which can deeply reduce the steric
hindrance. Two weak intermolecular hydrogen bonds are observed between the
oxygen atom from the central phenolic ring of one compound and the hydrogen atom
from the side phenolic ring of another one.
In the case of 3b, the OH groups from three phenolic rings are neighboring and the
oxygen atoms form a linear arrangement with two short separations of 2.7484(15) and
2.7451(16) Å between the central and outer oxygen atoms. Two intramolecular
hydrogen bonds have been produced due to the proximity of these phenolic groups,
resulting in a U-shaped configuration of the molecule. In the crystals, the side
phenolic rings from two different molecules generate significantly face-to-face π-π
stackings in an offset manner and the formed dimers are isolated from each other.
IR spectra
Fig. 3. Infrared spectra of the traditional antioxidants and the synthesized compounds.
The right picture is a large version of the figure in the left dotted box.
IR spectra of five phenols using KBr disc technique were provided in Fig. 3. The
major differences of the five phenols in IR spectra are the peaks with the absorption
above 3000 cm-1 which are the characteristic peaks of OH groups. In the infrared
spectra of BHT (1), significant signal is present at 3626 cm-1 attributed to the free OH

group. 2a containing two OH groups shows only a single sharp line characteristic of
free OH centered at 3641 cm-1, similarly to BHT. It is remarkably demonstrated that
the structure is symmetrical without any hydrogen interaction. However, 2b shows
two different hydroxyl absorptions including a sharp one (3601 cm-1) and a much
broader one (3393 cm-1) which can be concluded that one is free OH group and the
other is intramolecular hydrogen bonded, similarly to the reports [10]. In the infrared
spectra of trisphenol 3a, one signal at 3623 cm−1 and a shoulder peak at 3648 cm-1 are
the characteristic peaks of free OH. The other OH peak appears at 3557 cm-1 probably
due to the weak intermolecular hydrogen bonding. For the case of 3b, the peak at
3584 cm-1 can be assigned to the free OH and two other broad ones between 3300
cm-1 to 3400 cm-1 are the OH groups that participate in the intramolecular hydrogen
bonding. All the results are in a favorable consistence with that of the X-ray
diffractions.
3.2 Thermal analysis under nitrogen or air
The thermal stabilities of the phenols were then investigated under nitrogen or air
and the results are shown in Fig. 4 and Table 3. Ton and Tmax are defined as the
temperature at 5% weight loss and the temperature of maximum weight loss rate,
respectively; αmax is the maximum weight loss rate. The results of the thermal stability
investigations under inert conditions, presented in Fig. 4a and 4b, showed that all of
the antioxidants undergo weight loss over a single stage. The Ton and Tmax of BHT are
o
o
o
around 91 C and 138 C, respectively. Bisphenol 2a provides a higher Ton (182 C)
o
o
and Tmax (243 C) than that of 2b (177 and 239 C, respectively). Compared to the
bisphenols, the thermal stabilities of trisphenols make a large improvement which
provide a proximate Ton (245 oC) and Tmax (310 oC). It should be noted that 3a with a
higher molecular weight exhibits almost identical thermal stability to that of 3b. This
t
is probably because the more numbers of the Bu group in 3a’s structure would make
it easier to take off small molecules while the intramolecular hydrogen bonds in 3b
may partly hinder the release of water.

Fig. 4. TG and DTG curves of the traditional antioxidants and the synthesized
compounds: (a) The TG curves under nitrogen; (b) The DTG curves under nitrogen;
(c) The TG curves under air; (d) The DTG curves under air.
Table 3. Thermal decomposition data for the traditional antioxidants and the
synthesized compounds
Samples
1
2a
2b
3a
3b
Molecular weight (g/mol) 235.19 424.67 340.50 586.89 502.73
Ton (oC)
91
138
182
243
177
239
245
309
245
310
Nitrogen
Air
Tmax1 (oC)
αmax1 (%/oC) 2.852 2.495 2.206 2.376 1.768
Ton (oC)
Tmax1 (oC)
Tmax2 (oC)
88
135
-
184
244
-
175
228
527
240
273
523
237
284
559
αmax1 (%/oC) 2.994 2.811 1.987 0.651 0.553
αmax2 (%/oC)
-
-
0.185 0.333 0.495
The thermo-oxidative stability under reactive conditions was also investigated,
shown in Fig. 4c and 4d. Under air condition, 1 and 2a have a single-step degradation
progress but the other antioxidants have a two-step degradation progress which is
different from the results under nitrogen. The first decomposition process of 2b, 3a

o
and 3b occur between 200-300 C probably induced by easily releasing of water and
small molecules from the compounds. And at the higher temperature of this stage
some stable intermediates may possibly be produced through the reaction with oxygen,
o
leading to the second degradation step occurred at around 520-560 C. It is worth
noting that under air condition the trisphenols still behaved a much better
thermo-oxidative stability than bisphenols and BHT although the Ton and Tmax1
exhibited a little decline compared to nitrogen atmosphere. All the above results
demonstrate that the trisphenols are suitable to apply at elevated temperatures (>200
oC).
3.3 Antioxidant activity in lubricant base oil.
Fig. 5 Comparison of the antioxidant activities between bisphenols and trisphenols in
150N in different concentrations using isothermal PDSC.
In order to investigate the actual antioxidant performance, bisphenols and trisphenols
were added into lubricant base oils and their performance were tested using PDSC
[35-37], shown in Fig. 5. In order to accelerate the autoxidation, the samples were
o
heated at 180 C where the pressure of oxygen was kept on 3.5 MPa during the test.
The performance of compound 1 was not investigated due to its poor thermal stability
under the test conditions. The oxidation process can be monitored by the exothermic
situation. In the presence of antioxidants, the oxidation of base oil was inhibited
without any energy loss of the system. Once the antioxidants were exhausted, base oil
was quickly oxidized accompanied by the drastic heat release. Thus the antioxidant

activity can be evaluated by the oxidation induction time (OIT) which was measured
from the start of the oxygen added in to the start of the exothermic peak. On adding of
3 μmol/g different antioxidants, 3b exhibited the highest OIT value of 18.87 min,
followed by 18.73 min for 2b, 10.95 min for 2a and 8.81 min for 3a. Compared with
the pure oil, the OIT achieved a 5-fold increase for 2b and 3b, 2.6-fold increase for 2a
and 2-fold increase for 3a. Increasing the concentration of the antioxidants to 5
μmol/g afforded the OITs with a 7-fold increase for 2b and 3b, 3.9-fold increase for
2a and 3.3-fold increase for 3a compared with the pure oil. It is clearly indicated that
those compounds containing the intramolecular hydrogen bonds (2b and 3b) exhibited
much better performance than others and their antioxidant activity have a larger
improvement when increasing the concentration. Furthermore, it can be also found
that at lower concentration 3b performed a better activity than 2b, while at higher
concentration came out the opposite result. On the contrary, 3a always exhibited the
worst antioxidant activity of the four compounds.
Fig. 6 Comparison of the antioxidant activities between bisphenols and trisphenols in
150N in different concentrations using RBOT.
Except the high temperature and the pressure, metal ions and water are two of the
most common factors that can accelerate the oxidation of oils in practice. Thus their
antioxidant activities were also investigated in the presence of oxygen and catalyst
(water and copper) using RBOT and the results are shown in Fig 6. The RBOT time
of pure 150N is only 51 min. Adding of 3 μmol/g antioxidants can enhance the RBOT

time with a 4.2-fold increase for 3b, 3.5-fold increase for 2b, 1.3-fold increase for 2a
and 0.9-fold increase for 3a. Increasing the concentration of the antioxidants to 5
μmol/g can largely improve the antioxidant activity for the oils containing 2b and 3b,
but perform little influence for that of 2a and 3a. And the performance of 3b became
a little poorer than that of 2a, different from when the concentration is 3 μmol/g. In
general, these phenols exhibited poorer activities in the RBOT tests than the PDSC
tests which can be concluded that water and copper partly weaken the activity of the
phenols during the antioxidation process. Nevertheless, the trend of the antioxidant
activities has a good consistency with that of PDSC.
Fig. 7. Comparison of the antioxidant activities between bisphenols and trisphenols in
other lubricant oils using isothermal PDSC.
Their antioxidant activity was further investigated in different base oils using
isothermal PDSC. The oxidation induction times (OITs) containing 5 μmol/g of
various antioxidants are shown in Fig. 7. According to the American Petroleum
Institute (API) [38], 150SN belonging to group I has less than 90% (m/m) saturates
and more than 0.03% (m/m) sulfur. YUBASE (group III) are hydrogenated base oils
containing more saturates (≥ 90% m/m) and less sulfur (≤ 0.03% m/m). Poly alpha
olefin (PAO) is a kind of synthetic base oil, belonging to group IV. All of these
phenolic compounds have good solubility in the three types of base oils and the
solubility can be ranked as 150SN>YUBASE>PAO which can be concluded that a
lower level base oil is accompanied by a better solubility of the phenols. Bisphenols

show better solubility than trisphenols but they were totally dissolved during the
whole test. The OITs of the base oils themselves without any antioxidants are all less
than 4 min. In contrast, after adding phenols to these oils, the OITs achieved 2.6 to
5-fold increase for 2a, 4.3 to 8.7-fold increase for 2b, 2.0 to 5-fold increase for 3a and
4.2 to 8.0-fold increase for 3b. They both obtained the best antioxidant efficiency in
PAO but the worst in 150SN since more unsaturates that easier to autoxidize are
contained in 150SN. In all the base oils, the antioxidant activities can be ranked as
2b>3b>2a>3a, similarly to the results above.
3.4 DFT caculations
3.4.1 Calculations of BDE and IP
Table 4. Calculated BDE and IP values of the most stable isomers in the gas phase
and n-hexane.
Calculative content Solvent
2a
2b
3a
3b
Gas phase 76.16 74.94 75.61 75.30
n-hexane 77.43 74.27 76.98 73.89
Gas phase 168.30 165.97 166.44 162.40
n-hexane 154.51 149.39 154.04 146.13
BDE (kcal•mol-1)
IP (kcal•mol-1)
To study the radical scavenging activities of phenols as antioxidants, thermodynamic
properties consisting of bond dissociation enthalpy (BDE) and ionization potential (IP)
values in gas-phase and n-hexane (the solvent which was performed in the experiment)
were calculated and the results are shown in table 4. Fig. 8 presents the most stable
configurations of the formed phenolic radicals. In general, the highly effective
antioxidants have relatively low BDE and IP values. The BDE calculation is on the
base of a HAT mechanism which evaluates the easiness of the homolytic cleavage of
the bond [39]. In the gas phase two of the O-H groups in compound 2a are identical
with the same BDE values of 77.43 kcal•mol-1. The most active O-H group for 2b, 3a
and 3b is the free one without any inter- or intra-molecular hydrogen bonding and the
BDE values are 74.27 kcal•mol-1, 75.61 kcal•mol-1 and 73.89 kcal•mol-1 respectively.
In n-hexane, the BDE values of 2a and 3a increase with about 1.3 kcal•mol-1 while
the values of 2b and 3b decrease with 0.67 and 1.41 kcal•mol-1 respectively. The

increasing wide gaps of the BDE values of the four phenols obviously show that the
O-H groups of 2b and 3b become more active than that of 2a and 3a in nonpolar
solvent, verifying the better antioxidant performance of 2b and 3b in the lubricant oils
since most of the compositions in oils are hydrocarbon compounds.
Fig 8. The most stable configurations of the formed phenolic radicals.
The IP calculation undergoes the SET-PT mechanism including the process of
losing one electron to generate a radical cation and the following deprotonating of the
radical cation to create the corresponding radical [40]. Thus the IP values illustrate the
easiness of electron donation of phenolic compounds. In the gas phase, 3b shows the
lowest IP value of 162.4 kcal•mol-1, followed by 2a of 165.97 kcal•mol-1, 3a of
166.44 kcal•mol-1 and 2a of 168.30 kcal•mol-1. While in n-hexane, a decrease of about
12-14 kcal•mol-1 have been observed for the IP values of all the phenols which are
quite larger than the variations of BDE values, obviously demonstrating that the

electron transfer process is more sensitive to the solvent. Nevertheless, both of BDE
and IP values of 2b and 3b in gas-phase and n-hexane are lower than that of other
compounds which indicates that 2b and 3b have higher antioxidant activities.
3.4.2 Calculations of the TS structures and the energy barriers for the reaction of
the antioxidants and alkyl peroxy radicals
Scheme 4. Oxidation mechanism in oils.
Fig. 9. Calculated TS structures for the reaction between phenols and a methylperoxyl
radical.

During the oxidation of base oil (Scheme 4), alkyl peroxy radical (ROO•) was the
most pernicious radical which can quickly abstract hydrogen from hydrocabon
molecule to accelerate the degradation of base oil (eq 1) [41]. In this regard, phenols
acted as the radical scavengers donated hydrogen atoms to terminate ROO• to afford
hydroperoxides (ROOH) (eq 2) and the obtained phenoxy radicals can rapidly react
with another ROO• to afford non-radical compounds (eq 3). Thus the inherent
antioxidant efficacies of the phenols can be evaluated by the reactivity with peroxyl
radicals since it is the reaction that competes with the rate-determining step of
oxidation [42-43]. In consequence, the TS structures and the energy barriers of the
inhibited step were calculated by DFT methods, shown in Fig. 9 and Fig. 10. This step
undergo a proton coupled electron transfer (PCET) mechanism including proton
transfer perpendicularly to the phenolic ring, together with electron transfer from a
lone pair on the oxygen of phenol to the radical. Compound 2b and 3b provide the
lowest Gibbs free energy of this step with 13.52 and 13.87 kcal/mol respectively,
followed by 2a (16.85 kcal/mol) and 3a (16.98 kcal/mol). The TS structures show that
the less bulky of “CH2” substituents on the o-position of O-H group in 2b and 3b
make it easier for the approach of MeOO• to the hydroxyl hydrogen. Interestingly, in
the TS of the four compounds, a syn arrangement of phenols and methylperoxyl
radical is preferred, afforded an additional overlap between 2p lone pairs on the inner
O-atom of MeOO• and the π electros of the phenolic rings which can further stabilize
the transition state [44-46]. Furthermore, it should be noted that the stabilization of
the formed phenoxy radical has also an important influence on the activities [10, 47].
The calculations show that the most stable configurations of 2b• and 3b• can be
further stabilized by the remained intramolecular hydrogen bonding which can
additionally accelerate the rate of eq. 2 (Fig. 8). In short, the less bulky substituents of
the substrate, the additional stabilization of the transition state and the more
stabilization of the generated phenolic radical simultaneously give rise to the higher
antioxidant activity of compound 2a and 3b. These calculations provided a favorable
consistency with the experimental results.

Fig. 10. The calculated free energy barriers for the reaction between phenols and a
methylperoxyl radical.
4 Conclusions
We have successfully synthesized two types of trisphenols bridged by two “CH2”
groups in different ways. The X-ray diffractions and IR spectra showed that 3a has an
S-shaped configuration with intermolecular hydrogen bonds while 3b presents a
U-shaped configuration with two intramolecular hydrogen bonds. They both showed
better thermal stability than the traditional ones in nitrogen or air due to their higher
molecular weight. When adding them into lubricant oil, compound 3b exhibited the
best antioxidant activity, nearly identical to that of 2b, while 3a performed the worst.
The computational results showed that 3b provides the lowest BDE and IP values
while 3a gives the highest results. Furthermore, the calculations for the reaction with
alkyl peroxyl radicals showed that the superior performance of 3b is due to three
factors including the less bulky substituents of the substrate, the additional

stabilization of the transition state from the overlap between 2p lone pairs and the π
electrons, and the more stabilization of the generated phenolic radical ascribed to the
two strong intramolecular hydrogen bonds.
Acknowledgements
We thank Professor Shuzhou Li, School of Materials Science and Engineering
Nanyang Technological University in Singapore, for providing helpful idea and
thoughts about the computation. This work was supported by “Thousand Talents
Program”, National Natural Science Foundation of China (Grand No. 21606247),
China Postdoctoral Science Foundation (Grand No. 2016M590555), Zhejiang
Postdoctoral Sustentation Fund, China (Grand No. BSH1502163) and the Natural
Science Foundation of Ningbo (Grand No. 2016A610260).
References
[1] R. Czochara, J. Kusio, M. Symonowicz, G. Litwinienko, Ind. Eng. Chem. Res. 55
(2016) 9887.
[2] L. T. Zhang, G. X. Cai, W. M. J. Eli, Lubr. Sci. 25 (2013) 209.
[3] Kumar, A. Antioxidants Applications and Global Markets, BCC Research, Mar.
2015.
[4] W. A. Yehye, N. A. Rahman, A. Ariffin, S. B. A. Hamid, A. A. Alhadi, F. A.
Kadir, M. Yaeghoobi, Eur. J. Med. Chem. 101 (2015) 295.
[5] W. Wang, P. Kannan, J. Xue, K. Kannan, Environ. Res., 151 (2016) 339.
[6] Y. Bolbukh, P. Kuzema, V. Tertykh, I. Laguta, J. Therm. Anal. Calorim. 94 (2008)
727.
[7] R. Gensler, C. J. G. Plummer, H. H. Kausch, E. Kramer, J. R. Pauquet, H. Zweifel,
Polym. Degrad. Stab. 67 (2000) 195.
[8] A. Yano, S. Watanabe, Y. Miyazaki, M. Tsuchiya, Y. Yamamoto, Tribol Trans. 47
(2004) 111.
[9] S. Yu, J. X. Feng, T. Cai, S. G. Liu, Ind. Eng. Chem. Res. 56 (2017) 4196.
[10] R. Amorati, M. Lucarini, V. Mugnaini, G. F. Pedulli, J. Org. Chem. 68 (2003)
5198.

[11] M. Lucarini, G. F. Pedulli, L. Valgimigli, R. Amorati, J. Org. Chem. 66 (2001)
5456.
[12] R. Amorati, J. Zotova, A. Baschieri, L. Valgimigli, J. Org. Chem. 80 (2015)
10651.
[13] M. Ogata, M. Hoshi, K. Shimotohno, S. Urano, T. Endo, J. Am. Oil Chem. Soc.
74 (1997) 557.
[14] C. Zhao, Z. -Q. Liu, Biochimie. 93 (2011) 1755.
[15] S. Dikalov, T. Losik, J. L. Arbiser, Biochem. Pharmacol. 76 (2008) 589.
[16] Gaussian 09, Revision A.1, M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E.
Scuseria, M. A. Robb, J. R. Cheeseman, G. Scalmani, V. Barone, B. Mennucci, G. A.
Petersson, H. Nakatsuji, M. Caricato, X. Li, H. P. Hratchian, A. F. Izmaylov, J.Bloino,
G. Zheng, J. L. Sonnenberg, M. Hada, M. Ehara, K. Toyota, R. Fukuda, J. Hasegawa,
M. Ishida, T. Nakajima, Y. Honda, O. Kitao, H. Nakai, T. Vreven, J. A. Montgomery,
Jr., J. E. Peralta, F. Ogliaro, M. Bearpark, J. J. Heyd, E. Brothers,K. N. Kudin, V. N.
Staroverov, R. Kobayashi, J. Normand, K. Raghavachari, A. Rendell, J. C. Burant, S.
S. Iyengar, J.Tomasi, M. Cossi, N. Rega, J. M. Millam, M. Klene, J. E. Knox, J. B.
Cross, V. Bakken, C. Adamo, J. Jaramillo, R.Gomperts, R. E. Stratmann, O. Yazyev, A.
J. Austin, R. Cammi, C. Pomelli, J. W. Ochterski, R. L. Martin, K. Morokuma, V.G.
Zakrzewski, G. A. Voth, P. Salvador, J. J. Dannenberg, S. Dapprich, A. D. Daniels, Ö.
Farkas, J. B. Foresman, J. V. Ortiz,J. Cioslowski, and D. J. Fox, Gaussian, Inc.,
Wallingford CT, 2009.
[17] Gaussview Rev. 3.09, Windows version. Gaussian Inc., Pittsburgh.
[18] J. Tomasi, B. Mennucci, E. Cances, J. Mol. Struc. Theochem. 464 (1999) 211.
[19] J. S. Wright, E. R. Johnson, G. A. DiLabio, J. Am. Chem. Soc 123 (2001) 1173.
[20] J. E. Bartmess, J. Phys. Chem. 98 (1994) 6420.
[21] M. M. Bizarro, B. J. C. Cabral, R. M. B. Santos, J. A. M. Simões, PureAppl.
Chem. 71 (1999) 1249.
[22] W. D. Parker, J. Am. Chem. Soc. 114 (1992) 7458.
[23] E. Wilhelm, R. Battino, Chem. Rev. 73 (1977) 1.
[24] Z. Marković, J. Tošović, D. Milenković, S. Marković, Comput. Theor. Chem.

1077 (2016) 11.
[25] J. J. Fifen, M. Nsangou, Z. Dhaouadi, O. Motapon, N. Jaidane, Comput. Theor.
Chem. 966 (2011) 232.
[26] J. Rimarčík, V. Lukeš, E. Klein, M. Ilčin，J. Mol. Struc. (Theochem) 952 (2010)
25.
[27] A. D. Becke, J. Chem. Phys. 98 (1993) 5648
[28] P. J. Stephens, F. J. Devlin, C. F. Chabalowski, M. J. Frisch, J. Chem. Phys. 98
(1994) 11623
[29] P. C. Hariharan, J. A. Pople, Theor. Chim. Acta. 28 (1973) 213.
[30] K. Fukui, J Phys Chem 74 (1970) 4161.
[31] K. Fukui, Acc Chem Res. 14 (1981) 363.
[32] Y. Zhao, D. G. Truhlar, Theor. Chem. Account. 120 (2008) 215.
[33] A. D. McLean, G. S. Chandler, J. Chem. Phys. 72 (1980) 5639.
[34] Compound 2a and 2b has already been reported in previous literatures. We
discussed their crystal structures here in order to better understand the relationship
between structures and properties of different phenols. For 2a, please see: T. J. Boyle,
L. A. M. Steele, D. T. Yonemoto, J. Coord. Chem. 65 (2012) 487. For 2b, please see:
L. A. Chetkina, V. E. Zavodnik, V. K. Bel'skii, I. G. Arzamanova, M. I. Naiman, and
Ya. A. Gurvich, J. Struct. Chem. 25 (1985) 935.
[35] C. Qiu, S. Han, X. Cheng, T. Ren, Thermochimica Acta 447 (2006) 36.
[36] C. Miao, Y. Zhang, G. Yang, S. Zhang, L. Yu, P. Zhang, Ind. Eng. Chem. Res.
55 (2016) 12703.
[37] C. Miao, D. Yu, L. Huang, S. Zhang, L.Yu, P. Zhang, Ind. Eng. Chem. Res. 55
(2016) 1819.
[38] Engine Oil Licensing and Certification System, 17th ed.; API Publication 1509;
American Petroleum Institute, September 2015.
[39] Y. Zhao, N. E. Schultz, D. G. Truhlar, J. Chem. Phys. 123 (2005) 161103.
[40] Z. Marković, D. Amić, D. Milenković, J. M. D. Marković, S. Marković, Phys.
Chem. Chem. Phys. 15 (2013) 7370.

[41] J. Dong, C. A. Migdal, Antioxidants. In Lubricant Additives: Chemistry and
Application, 2nd ed; Rudnick, L. R. Ed.; CRC Press: Boca Raton, 2009; pp 3-50.
[42] B. Li, J. R. Harjani, N. S. Cormier, H. Madarati, J. Atkinson, G. Cosa, D. A. Pratt,
J. Am. Chem. Soc. 135 (2013) 1394.
[43] G. W. Burton, K. U. Ingold, Acc. Chem. Res. 19 (1986) 194.
[44] G. A. DiLabio, K. U. Ingold, J. Am. Chem. Soc. 127 (2005) 6693.
[45] G. A. DiLabio, E. R. Johnson, J. Am. Chem. Soc. 129 (2007) 6199.
[46] L. Valgimigli, R. Amorati, S. Petrucci, G. F. Pedulli, D. Hu, J. J. Hanthorn, D. A.
Pratt, Angew. Chem., Int. Ed. 48 (2009) 8348.
[47] P. J. O’Malley, J. Phys. Chem. B 106 (2002) 12331.

Highlights
 Synthesis and characterization of two new type trisphenols
 Superior thermalstability and highly antioxidant efficiency of the synthesized
compounds in lubricant oils compared to the related phenols
 The relationship between structure and properties explored by a series of DFT
calculations