Experimental and computational studies on the formation and biological properties of the simplest polyfluoroalkyl phosphonates

Article abstract of DOI:10.1039/d0nj00963f

A representative set of the simplest dialkyl polyfluoroalkylphosphonates was in silico profiled against their absorption, distribution, metabolism, and excretion with the use of the SwissADME tool. Promising results of the screening led us to attempt to synthesize the title compounds for further biological investigations. Detailed experimental and quantum-theoretical (DFT) investigations were performed to reveal that the studied compounds could not be obtained through the standard Michaelis-Arbuzov or Michaelis-Becker reactions. Kinetic studies showed that trimethyl phosphite undergoes a reaction with 1,1,1-trifluoro-2-iodoethane several orders of magnitude slower than a side reaction. The difficulty was overcome by developing a simple three-step synthesis path that involves only readily available substrates. The newly synthesized substances were tested against several cell lines. The in vitro research revealed that both dimethyl and diethyl (2,2,3,3,3-pentafluoropropyl)phosphonate exhibited toxicity towards glioblastoma cells (U-87 MG) at a considerably lower concentration than known chemotherapeutics. This journal is

Full text of DOI:10.1039/d0nj00963f

View Article Online
View Journal
NJC
New Journal of Chemistry
Accepted Manuscript
A journal for new directions in chemistry
This article can be cited before page numbers have been issued, to do this please use: P. M. Zagórski, P.
Tokarz, B. Gostynski and P. Tokarz, New J. Chem., 2020, DOI: 10.1039/D0NJ00963F.
Volume 42
This is an Accepted Manuscript, which has been through the
Number
6
21 March 2018
Pages 3963-4776
Royal Society of Chemistry peer review process and has been
accepted for publication.
NJC
New Journal of Chemistry
rsc.li/njc
A journal for new directions in chemistry
Accepted Manuscripts are published online shortly after acceptance,
before technical editing, formatting and proof reading. Using this free
service, authors can make their results available to the community, in
citable form, before we publish the edited article. We will replace this
Accepted Manuscript with the edited and formatted Advance Article as
soon as it is available.
You can find more information about Accepted Manuscripts in the
Information for Authors.
Please note that technical editing may introduce minor changes to the
text and/or graphics, which may alter content. The journal’s standard
Terms & Conditions and the Ethical guidelines still apply. In no event
shall the Royal Society of Chemistry be held responsible for any errors
or omissions in this Accepted Manuscript or any consequences arising
from the use of any information it contains.
ISSN 1144-2546
PAPER
Chechia Hu, Tzu-Jen Lin et al.
Yellowish and blue luminescent graphene oxide quantum dots
prepared via a microwave-assisted hydrothermal route using
H2O2 and KMnO4 as oxidizing agents
rsc.li/njc

Page 1 of 24
New Journal of Chemistry
1
2
3
4
5
6
7
View Article Online
Experimental and computational studies on the formation and
DOI: 10.1039/D0NJ00963F
biological properties of the simplest polyfluoroalkyl phosphonates
Piotr M. Zagórski1, Paulina Tokarz2, Bartłomiej Gostyński3, Paweł Tokarz4
8
9
1University of Lodz, Faculty of Chemistry, Department of Organic Chemistry, Tamka 12, 91-
403 Lodz
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
2University of Lodz, Faculty of Biology and Environmental Protection, Department of
Molecular Genetics, Pomorska 141/143, 90-236 Lodz, Poland
3Centre of Molecular and Macromolecular Studies, Polish Academy of Sciences, Sienkiewicza
112, Lodz 90-363, Poland
4University of Lodz, Faculty of Chemistry, Department of Organic Chemistry, Molecular
Spectroscopy Laboratory, Tamka 12, 91-403 Lodz
Abstract
A representative set of the simplest dialkyl polyfluoroalkylphosphonates was in silico profiled
against their absorption, distribution, metabolism, and excretion with the use of the
SwissADME tool. Promising results of the screening led us to attempt to synthesize the title
compounds for further biological investigations. Detailed experimental and quantum-
theoretical (DFT) investigations were performed to reveal that the studied compounds could
not be obtained through the standard Michaelis-Arbuzov or Michaelis-Becker reactions. Kinetic
studies showed that the trimethyl phosphite undergoes a reaction with 1,1,1-trifluoro-2-
iodoethane at several orders of magnitude slower than a side reaction. The difficulty was
overcome by developing a simple three-step synthesis path that involves only readily available
substrates. The newly synthesized substances were tested against several cell lines. The in vitro
research revealed that both dimethyl and diethyl (2,2,3,3,3-pentafluoropropyl)phosphonate
exhibited toxicity towards the glioblastoma cells (U-87 MG) at a considerably lower
concentration than known chemotherapeutics.
Keywords: polyfluoroalkyl phosphonates, toxicity, glioblastoma, ADME, kinetics
Corresponding author: Paweł Tokarz; pawel.tokarz@chemia.uni.lodz.pl; +48 42 635 58 12
This work is dedicated to the memory of Professor Jarosław Lewkowski, our colleague, teacher and friend.
Additional data are available in the Electronic supporting information

New Journal of Chemistry
Page 2 of 24
1
2
3
4
5
View Article Online
DOI: 10.1039/D0NJ00963F
1. Introduction
The interest in a synthesis of fluorinated and perfluoroalkylated compounds is a still-growing
trend in chemistry. This is motivated by the fact that fluorine atoms notably change the
physicochemical properties of organic molecules, including their biological activity.
Particularly, perfluoroalkylated chains increase the lipophilicity and interactions with several
cellular structures.1 For these reasons trifluoromethyl group, which constitutes the smallest
polyfluorinated moiety that can be incorporated into a molecule in numerous ways, is
commonly found in medicines.2-7 Equally interesting are biochemical characteristics of the
phosphonic moiety. Phosphonates occur naturally in numerous structures of living organisms
and play an important role in several biochemical pathways.8 Specifically, they can be
recognized by enzymes related to phosphorylation/dephosphorylation processes and support
trans-membrane transport, particularly anionic chemical delivery system (aCDS), which is used
to deliver molecules to nerve cells.9, 10 This led us to a question of possible biological activity
of a molecule with both polyfluoroalkyl and phosphonic moieties combined with merely no
other functional group. Surprisingly, the literature study returned virtually no data related to
any biological activity of such structures – a strikingly peculiar gap when the molecular
simplicity of the compounds is considered.
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Not only the references to the bioactivity but also the synthesis of the simplest possible
polyfluoroalkyl phosphonates are limited. For the first time the dimethyl (2,2,2-
trifluoroethyl)phosphonate
was
obtained
by
the
hydrolysis
of
dimethyl
pentafluoroisopropenylphosphonate with only 10% yield.11 The first practically useful method
of the synthesis of the target compounds was developed in 2016 by the Gouverneur group.12
The method exploits 2,2,2-trifluoro-1-diazoethane which has been considered possibly
explosive.13 Recently, several dialkyl (2,2,2-trifluoroethyl)phosphonates have been prepared
from 1,1-dichloro-2,2,2-trifluoroethane or trifluoroacetaldehyde N-tosylhydrazone utilizing
copper-mediated synthesis.14, 15 While both, 2,2,2-trifluoro-1-diazoethane and 1,1-dichloro-
2,2,2-trifluoroethane are readily available from commercial sources they remain rather
expensive starting materials. Moreover, their homologs with longer fluorinated chains are less
accessible and even more expensive. Thus, we faced a problem of the accessibility of the desired
compounds.
Throughout the article, the dialkyl polyfluoroalkyl phosphonates of formula
CxFy(CH2)nP(O)(OR)2 (x=1,2, y=2x+1, n=1,2,3, R=Me, Et) will be referred to as ZOTy-n-R,
according to the naming convention used in our research project (Scheme 1).
Scheme 1. The studied compounds.
2. Results and discussion
2.1. ADME profiling
Due to a lack of biological data related to the ZOTs, we decided to perform in silico screening,
concerning the plausible absorption, distribution, metabolism, and excretion (ADME)
2

Page 3 of 24
New Journal of Chemistry
1
2
3
4
5
View Article Online
DOI: 10.1039/D0NJ00963F
processes descriptors. The analysis was performed using the SwissADME – a free-to-use tool
provided online by the Swiss Institute for Bioinformatics.16, 17
6
7
8
9
The tool provides results of bioinformatic calculations of several molecular descriptors
associated with physicochemical properties that strongly impact the biochemical properties of
a substance, including lipophilicity, size, polarity, insolubility, insaturation, and flexibility. We
tested trifluorinated and pentafluorinated phosphonates. Lipophilicity was calculated as
XLOGP3, an atomistic method including corrective factors and knowledge-based library.18 For
polarity, the tool uses a topological polar surface, an additive model based on contributions of
separate atoms considered polar.19 Solubility was predicted by the use of the ESOL descriptor
(Estimated SOLubility).20 Flexibility is simply expressed as the number of rotatable bonds. The
compound is considered promising if its descriptor values lie in the subspace of the
physicochemical properties suitable for bioavailability. Graphically, the bioavailability
subspace can be represented by an area on a six-axis plot on which the most critical parameters
are taken into account. It is clearly visible that considered fluorinated phosphonates fit perfectly
into the desired subspace (Fig. 1).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 1. Radar plots of the main ADME parameters: lipophilicity (XLOGP3), molecular size,
polarity (TPSA), insolubility (log S(ESOL)), insaturation (fraction of the sp3 carbons) and
flexibility (number of rotatable bonds). The grey area represents the suitable physicochemical
subspace for oral bioavailability according to the literature.16 Consult the main text (Section
2.1) and the corresponding Table 1 for further details and values.
The selected descriptors critical for predicting the pharmacokinetics and the druglikeness of the
compounds have been analyzed and summarized in Tab. 1. All tested ZOTs showed high
gastrointestinal(GI) absorption and blood-brain barrier (BBB) permeability.21 Additionally, all
tested ZOTs were estimated to be non-substrates of p-glycoprotein (P-gp) and non-inhibitors
of a given cytochrome P450 (CYP) isomers using algorithm design in SwissADME.17 P-gp is
a key member of ATP-binding cassette transporters, which protects the central nervous system
3

New Journal of Chemistry
Page 4 of 24
1
2
3
4
5
6
7
8
9
View Article Online
(CNS) from xenobiotics and is associated with multidrug-resistance (MDR) in cancers. CYP
isomers are the major executors of xenobiotics biotransformation and the alteraDtiOoI:n10o.10f39th/De0iNrJ00963F
activity can result in harmful drug-drug interactions due to the production of metabolites with
the pharmacological and toxicological profiles different from the original drug. Since ZOTs are
predicted to be inert to P-gp and CYP isomers it can be expected that ZOTs should be resistant
to P-gp-mediated drug elimination and its intake should be safe in the respect of drug
metabolism.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
The Lipinski22, Ghose23, Veber24, Egan25, and Muegge26 rules were applied to evaluate
druglikeness to predict whether a compound is likely to be orally bioavailable. The number of
violations to these rules is documented in Tab. 1. All analyzed compounds lie within the range
of all the parameters of the Lipinski, Veber, and Egan rules and demonstrated the bioavailability
score of 0.55. According to the evaluation, ZOT3-1-Me, ZOT5-1-Me, ZOT3-2-Me lie within
the close agreement with the Ghose and Muegge rules while the rest of the compounds abide
with the criteria.
Table 1. The main results of the ADME analysis of ZOTs. The minimum and maximum values
describe the subspace for oral bioavailability. For a detailed description consult the main text
(Section 2.1). The full set of data returned by SwissADME is available in the Supplementary
Material file.
Molecule
ZOT3-1-Me ZOT3-1-Et ZOT5-1-Me ZOT5-1-Et ZOT3-2-Me ZOT3-3-Me Min Max
150 500
0,25 1,00
MW
Insaturation
192,07
1
220,13
1
242,08
1
270,13
1
206,1
1
220,13
1
0
9
Rotatable bonds
Polarity TPSA
Lipophilicity XLOGP3
Insolubility ESOL Log S
GI absorption
4
6
5
7
5
6
20 130
0,70 50
-6
45,34
0,59
-1,14
High
Yes
No
No
No
No
No
No
0
45,34
1,32
-1,64
High
Yes
No
No
No
No
No
No
0
45,34
1,26
-1,8
High
Yes
No
No
No
No
No
No
0
45,34
1,99
-2,31
High
Yes
No
No
No
No
No
No
0
45,34
0,94
-1,38
High
Yes
No
No
No
No
No
No
0
45,34
1,3
-1,63
High
Yes
No
No
No
No
No
No
0
0
BBB permeant
P-gp substrate
CYP1A2 inhibitor
CYP2C19 inhibitor
CYP2C9 inhibitor
CYP2D6 inhibitor
CYP3A4 inhibitor
Lipinski #violations
Ghose #violations
Veber #violations
Egan #violations
Muegge #violations
Bioavailability Score
PAINS #alerts
2
0
0
2
0
0
0
0
1
0
0
0
0
0
0
0
1
0
0
0
0
0
0
0
0,55
0
0,55
0
0,55
0
0,55
0
0,55
0
0,55
0
Leadlikeness #violations
MW – molecular weight
1
1
1
0
1
1
GI absorption – gastrointestinal absorption
BBB – blood-brain barrier permeability
P-gp – p-glycoprotein
SwissADME performs an analysis of pan assay interference compounds (PAINS)27 which allow
to exclude compounds that give the false-positive response in biological assays from the drug-
discovery process. The evaluation of ZOTs returned no warnings with the PAINS filter. All
remaining data are available in the supplementary material.
4

Page 5 of 24
New Journal of Chemistry
1
2
3
4
View Article Online
DOI: 10.1039/D0NJ00963F
2.2.Michaelis-Arbuzov reaction kinetics
5
6
7
8
Promising ADME parameters fully justified the attempts of the synthesis of ZOTs. We decided
to keep the synthesis as simple as possible exploiting only readily available substrates. This is
essential when higher fluorinated homologs are considered. This is why we initially focused on
the Michaelis-Arbuzov protocol. The reaction alkyl halide and trialkyl phosphite is a standard
method of choice when the synthesis of an alkylphosphonates is desired. Thus, our first attempt
to obtain the polyfluoroalkyl phosphonates was to exploit a set of reactions between the
appropriate polyfluoroalkyl iodides and trimethyl phosphite (TMP) (Scheme 2, reaction 1).
We expected a formation of minor amounts of 2 as a typical by-product – a result of the side
reaction between the trimethyl phosphite and methyl iodide liberated during the main reaction
(Scheme 2, reaction 2). Surprisingly, the 1H and 31P NMR spectra recorded from the reaction
mixture did not reveal the formation of the desired fluorinated phosphonate at all. Instead, the
only detectable product observed was dimethyl methyl phosphonate (2). Regardless of the
conditions applied, the only process that took place was the formal rearrangement of the TMP
to 2.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Scheme 2. The desired Michaelis-Arbuzov reaction (1) and the side-reaction (2) of methyl
iodide.
We hypothesized that reaction 1 is of several orders of magnitude slower than reaction 2 thus
the amount of ZOT produced is too low for being observed on NMR while the amount of
methyl iodide being enough to sustain the rearrangement cycle. Typically, one could simply
perform the reaction in an open flask to allow the methyl iodide to leave the system. However,
the volatility of the substrate rendered such technique inapplicable. We attempted to verify this
hypothesis by kinetic experiments. Since direct measurement of the kinetics of the reaction 1
would be virtually impossible due to an extremally low rate of the process we developed a
mathematical model of the reaction system from which the desired parameters could be
extracted. The dependency of the concentration of 2 on the reaction time found with aid of the
Wolfram Development Platform at Wolfram Cloud28 follows the formula: (Eq. 1):
2
√
푥푀푒푃 푂
= 푥0,푃 푂푀푒 3 tanh (푡 1 푘1푘2푥퐶퐹 (퐶퐻 ) 퐼 푥0,푃 푂푀푒
)
( 1 )
(
)(
푂푀푒
)
(
)
(
)
2
3
2
푛
3
2
To estimate the reaction rate constants k1 and k2 we performed two series of NMR-controlled
kinetic experiments. First, the kinetics of the simpler reaction 2 was examined by heating the
samples of trimethyl phosphite containing respectively 0, 0.01, 0.02, and 0.04 mole fraction of
methyl iodide. The reactions were run at 50 °C and the progress was monitored by 31P and 1H
NMR for 16 hours (see the Experimental section for technical details).
The initial reaction rate v0 is linearly proportional to the reaction rate constant k2 and the
concentration of methyl iodide. This is understandable, as methyl iodide can be formally
considered a catalyst for the reaction.
푣
0
= 푘2푥푀푒퐼
( 2 )
푥
(
)
0,푃 푂푀푒
3
5

New Journal of Chemistry
Page 6 of 24
1
2
3
4
5
View Article Online
The value of v0 for each methyl iodide concentration was estimated from the initial data points
as presented in Fig. 2. The reaction rate constant calculated by regression was fouDnOdI:t1o0.1e0q39u/Da0tNeJ00963F
k2 = 7.24 ∙ 10-4 s-1.
6
7
8
9
To determine the rate constants k1 of the reaction 1 for each of three different fluorinated iodides
(n=1,2,3) three separate reaction mixtures with trimethyl phosphite were prepared. In each of
them, the molar fraction of the fluorinated iodide was equal to 푥퐶퐹 (퐶퐻 ) 퐼 = 0.1 (푥0,푃 푂푀푒
=
)
3
(
3
2
푛
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
0.9). The reactions were run at 50 °C and monitored on 31P and 1H NMR.
Figure 2. Left: the concentration of the product of the Michaelis-Arbuzov reaction vs. the
reaction time depending on the methyl iodide concentration. Right: dependence of the initial
reaction rate vs. the methyl iodide concentration.
A qualitative examination of the plotted data showed the expected “sigmoid” shape (Fig. 3).
However, the attempt of fitting the squared hyperbolic tangent function in accordance with Eq.
1 to the data revealed a significant discrepancy between the theoretical model and the empirical
results. As a plausible explanation to this could serve the fact that the theoretical model does
not account for the changes of the physical properties, especially dielectric constant and
viscosity of the reaction medium throughout the reaction progress (at the beginning the reaction
mixture is dominated by the trimethyl phosphite which becomes successively replaced by the
phosphonate 2). The incorporation of the changes of the reaction medium properties into a
solvable differential equation system would require unsolicited assumptions that we decided to
avoid. However, during the initial stage of the reaction progress, it can be safely assumed that
푥푃 푂푀푒 3and the dielectric constant remains quasi-constant. In such case the solution to the
(
)
differential equation system simplifies to becomes quadratic (Eq. 3):
1
푥푀푒푃 푂
= 푡푘1푘2(푥0,푃 푂푀푒 )2푥퐶퐹 (퐶퐻 ) 퐼
( 3 )
(
)(
푂푀푒
)
(
)
3
2
3
2 푛
2
This should be expected, as the initial increase of the concentration of the methyl iodide
produced in reaction 1 is linear thus the increase of the concentration of the product of the
reaction 2 catalyzed by this iodide must be a quadratic function of the reaction time. After fitting
the data from the initial stage of the reaction progress (arbitrarily set to 푥푀푒푃 푂
≤ 0.2)
)( )
푂푀푒
2
(
with a quadratic function, the k2 values for each substrate were easily calculated from the
second-order coefficient of the resulting trinomial (Fig. 3).
6

Page 7 of 24
New Journal of Chemistry
1
2
3
4
View Article Online
DOI: 10.1039/D0NJ00963F
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 3. Reaction progress as measured by the integration of 31P NMR spectra. Top: the full
set of data with the data used for fitting marked as green; note the sigmoidal shape of the curve,
reassembling the squared hyperbolic tangent, as predicted by Eq. 1. Bottom: magnified part of
the data taken for fitting the quadratic function as predicted by Eq. 3. Black line – the fitting
curve.
The values of the calculated reaction rate constants are presented in the Tab. 2. It is clearly
visible that theArbuzov reaction of the 1,1,1-trifluoro-2-iodoethane is about 350 thousand times
slower than the side reaction with the methyl iodide. Even in the case of 1,1,1-trifluoro-4-
iodobutane, the main reaction is almost 3000 times slower than the side reaction. Elongation of
the linker between the CF3 group and the iodine atom by one methylene moiety causes the
reaction to speed up about ten times.
Table 2. Reaction rate constants of the Arbuzov reaction for different iodides.
Substrate
Reaction rate constant / s-1
7.24 ∙ 10-4
CH3I
CF3CH2I (1a)
CF3CH2CH2I (1b)
CF3CH2CH2CH2I (1c)
2.06 ∙ 10-9
2.46 ∙ 10-8
2.55 ∙ 10-7
2.3.Michalis-Becker reaction
To avoid the above described competitive side reaction of methyl iodide liberated from TMP,
we attempted to employ the Michaelis-Becker protocol instead. The appropriate anionic
phosphonic nucleophile was generated from dimethyl H-phosphonate and sodium hydride in
DMF solution (Scheme 3, reaction 3) and treated with an appropriate iodide (1a-c) at room
temperature (Scheme 3, reaction 4).
7

New Journal of Chemistry
Page 8 of 24
1
2
3
4
View Article Online
DOI: 10.1039/D0NJ00963F
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Scheme 3. Michaelis-Becker synthesis: generation of the phosphite salt (3) and its reaction
with the appropriate iodide.
This approach allowed us to obtain products ZOT3-2-Me and ZOT3-3-Me with yields of 16%
and 62% respectively (Tab. 3). However, no traces of ZOT3-1-Me have been observed. In all
cases, the formation of byproduct 2 was observed. However, its origin is different than in the
case of the Michaelis-Arbuzov reaction – it is probably formed in a nucleophilic attack of the
anion 3 on the methoxy group from dimethyl H-phosphonate (Scheme 4). Prolonged reaction
time did not affect the yield distribution significantly.
Scheme 4. The plausible mechanism of the side reaction of the Michaelis-Becker synthesis in
which the 2 is formed along with the desired fluorinated phosphonate.
Table 3. Yields (isolated) of the Michaelis-Becker reaction for different iodides.
Substrate
Yield / %
0
CF3CH2I (1a)
CF3CH2CH2I (1b)
CF3CH2CH2CH2I (1c)
16.1
62.2
2.4.Theoretical calculations
Comparing the results obtained from the Michaelis-Becker and Michaelis-Arbuzov reaction it
was clearly visible that the proximity of the trifluoromethyl group causes a significant decrease
in the reactivity of the electrophilic center in fluorinated iodides.
For a better insight into the molecular course of the substitution, we performed DFT
calculations. First, the Michaelis-Becker reaction was chosen for modeling as its mechanism is
simpler. All reactions were modeled as solutions in DMF (consult the Experimental section for
technical details). The overall process was modeled as two separated reactions – formation of
the sodium salt and the substitution of the iodide.
The first stage (Scheme 3, reaction 3) is the same in all cases thus it does not contribute to the
observed differences. The reaction runs in an open-system and the newly-created hydrogen gas
leaves the medium.
8

Page 9 of 24
New Journal of Chemistry
1
2
3
4
5
View Article Online
Table 4. Energetic contributions of reactions 3-4 in the total reaction referring to reactions
occurring in DMF solution (temperature 273.15 K for reaction 3 and 298.15 K forDrOeI:a1c0.t1i0o39n/D40)N.J00963F
6
7
8
9
Reaction 3
ΔH / kcal mol-1
-23.98
ΔG / kcal mol-1
-23.07
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Reaction 4
Substrate
ΔH / kcal mol-1
-51.12
ΔG / kcal mol-1
-49.36
CF3CH2I (1a)
CF3CH2CH2I (1b)
CF3CH2CH2CH2I (1c)
-51.41
-49.43
-50.41
-48.06
The second reaction (Scheme 3, reaction 4) was considered the limiting step thus we discuss
it in detail. Free energy changes for n=1,2,3 indicate that in each case the process is
thermodynamically preferred (Tab. 4), thus differences in the reaction yields may exist due to
kinetic factors.
Figure 4. TS structure for the proposed SN2 mechanism of reaction 2. The image represents
the case n=1, for n=2,3 analogously.
The considered mechanism of the reaction is an SN2 one, where the lone pair of the phosphorus
atom is engaged in a nucleophilic attack on the carbon center linked to iodine atom. Unlike the
typical mechanism for SN2 cases, transition state (TS) eventually found by us using the QST2
method represents a side-attack approach, not a backside one (Fig. 4).
At first glance, the calculated differences in total reaction energy provide almost no explanation
for the high discrepancies in the reactivity among the substrates. Thus, to analyze the kinetics
of the reactions the free energies of activation for each system were calculated as the differences
(Eq. 4):
Δ퐺푛‡ = 퐺푛,푇푆 − 퐺푛,퐶푂푀푃퐿
,
(4)
where GTS and GCOMPL are the values of free energy for transition state and substrate complex
respectively. Next, according to the Arrhenius equation (Eq. 5), the kinetic rate constants kn
have been expressed as:
‡
푛
−Δ퐺
푅푇
푘푛 = 퐴푛e
,
(5)
9

New Journal of Chemistry
Page 10 of 24
1
2
3
4
5
6
7
8
View Article Online
where k is the rate constant of considered reaction, A is an empirical value, the so-DcOaIl:l1e0d.10p39r/eD-0NJ00963F
exponential factor, R – the universal gas constant and T – temperature. After a reasonable
assumption of constancy of A for n = 1,2,3 in the given reaction conditions, Arrhenius
equation (Eq. 6) may serve to designate the quotients of individual rate constants k1, k2 and k3
respectively:
9
‡
푖
푅푇
‡
푛
Δ퐺 −Δ퐺
푘
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
푛
= e
.
(6)
푘
푖
A comparison of the kn/k3 rate constant quotients with observed reaction yields is particularly
informative (Tab. 5). Although all thermodynamically preferred, the individual reactions do not
proceed with the same (or even similar) rate. The substrate 1c reacts at least 2 orders of
magnitude faster than 1a and only about twice as fast as 1b. This perfectly corresponds to the
isolation yields.
Table 5. Free energies of activation and relative rate constants of the Michaelis-Becker reaction
at 273.15 K. Values in parentheses refer to 298.15 K.
Substrate
ΔG‡ / kcal mol-1
39.48 (39.52)
36.06 (36.13)
35.59 (35.60)
kn/k3
CF3CH2I (1a)
0.0008 (0.0033)
0.4175 (0.4041)
1 (1)
CF3CH2CH2I (1b)
CF3CH2CH2CH2I (1c)
To gain insight into the obtained results we studied in silico the Michaelis-Arbuzov reaction.
For the sake of comparison, besides the fluorinated alkyl iodides (R = CF3(CH2)n-1) we decided
to calculate the values of thermochemical potentials (ΔH, ΔG) for the reaction of methyl iodide
(R = H) which, according to the experimental results, proceeds several orders of magnitude
faster.
According to the accepted SN2-mechanism, an alkyl halide undergoes a nucleophilic attack by
the phosphorus lone pair of a trivalent phosphorus ester, yielding a quaternary phosphorus salt.
In the subsequent step, a nucleophilic attack of the counter anion on the alkoxy-substituent of
the salt creates another alkyl halide and a dialkyl phosphonate (Scheme 5).
The PES stationary points for Michaelis-Arbuzov reaction, especially these corresponding to
the transition state of the back-side nucleophilic attack of the phosphorus lone pair on the alkyl
iodide, were obtained easily. Energetic results for the reaction path were obtained for 273.15 as
well as 298.25 K (Tab. 6).
The listed values show that energetic outcomes for reactions 3a-b exhibit practically no
temperature dependence, and their values are very close, which means that all observed
discrepancies in the reaction rates should be due to kinetic factors, similarly as in the case of
reaction 2.
10

Page 11 of 24
New Journal of Chemistry
1
2
3
4
View Article Online
DOI: 10.1039/D0NJ00963F
5
6
7
8
R = H:
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
R = CF3:
Scheme 5. The plausible mechanism of the Michaelis-Arbuzov reaction (R = H or CF3(CH2)n-1)
and the calculated transition states TS1 and TS2 for two cases.
Table 6. Values of overall changes in thermochemical potentials for the Michaelis-Arbuzov
reaction in DMF solution in temperature 273.15 K as well as 298.15 K (parentheses). All values
in kcal/mol.
Substrate
ΔH / kcal mol-1
-38.59 (-38.60)
-37.69 (-37.69)
-37.96 (-37.97)
-36.97 (-36.98)
ΔG / kcal mol-1
-37.32 (-37.20)
-36.21 (-36.07)
-36.29 (-36.14)
-34.95 (-34.77)
MeI
CF3CH2I (1a)
CF3CH2CH2I (1b)
CF3CH2CH2CH2I (1c)
Two PES stationary points correspond to two transition states: TS1, connected to the
nucleophilic attack of phosphorus lone pair, and TS2, connected to the attack of the counterion.
‡
Relative free energy barriers associated with them and depicted in Tab. 6 are denoted as ΔGTS1
and ΔGTS2‡ respectively and calculated as:
Δ퐺푛‡푇푆1 = 퐺푛,푇푆1 − 퐺푛,퐶푂푀푃퐿
for TS1 and
Δ퐺푛‡,푇푆2 = 퐺푛,푇푆2 − 퐺푛,푆퐴퐿푇
for TS2, where Gn,TS1 and Gn,TS2 are free energy values for corresponding transition states while
Gn,COMPL and Gn,SALT are free energy values for substrate complex and quaternary salt
respectively. As follows from the presented outcomes – the nucleophilic attack of the
phosphorus lone pair is the rate-limiting step of the reaction considered.
11

New Journal of Chemistry
Page 12 of 24
1
2
3
4
5
6
7
8
9
View Article Online
Table 7. Energy barrier values for both transition states of Arbuzov reactiDoOnI: 10a.1n0d39/Dth0NeJ00963F
corresponding calculated and experimental rate constant quotients at 273.15 K. Values in
parentheses refer to 298.15 K, experimental values refer to 323.15 K.
Δ퐺푛‡푇푆1
/ kcal mol-1
Δ퐺푛‡푇푆2
/ kcal mol-1
Substrate
kn/kMeI; calc.
kn/kMeI; exper.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
MeI
1a
16.71 (16.91)
25.26 (25.45)
21.85 (22.16)
18.02 (18.04)
14.51 (14.49)
14.65 (14.73)
13.44 (13.44)
13.01 (13.02)
1.0
(1.0)
1.0
1.4·10-7
7.6·10-5
8.9·10-2
(5.3·10-7)
(1.4·10-4)
(1.4·10-1)
2.84·10-6
3.40·10-5
3.51·10-4
1b
1c
It is visible that the general trends observed in the experiment are closely followed by the
theoretical results except for the compound 1c. Nevertheless, in the case of 1c, the observed
discrepancy of two orders of magnitude translates to roughly 2.0 kcal/mol error in
Δ퐺3‡푇푆1value – the applied level of theory permits such an error. Moreover, the longer the
fluoroalkyl chain, the more conformers (i.e. local minima on PES) it possesses. Apparently,
G3,COMPL-minimum found by us (and used in the Δ퐺3‡푇푆1calculations) is not a global minimum
but a local one that is about 2.0 kcal/mol higher in energy than the global one. The reaction rate
heavily depends on the proximity of the CF3 group. Substrate 1a reacts about 6-7 orders of
magnitude slower than MeI.
2.5.Synthesis
Scheme 6. The three-step synthesis of the ZOT3/5-1 compounds.
Since the Arbuzov reaction proved to be unsuccessful in the case of compounds with the
shortest carbon chain we developed another strategy to synthesize the desired fluorinated
phosphonates(Scheme 6). The protocol started from relatively cheap and readily available
perfluorinated aldehyde hydrates (either fluoral hydrate or pentafluoropropionaldehyde
hydrate). Upon treatment with dialkyl H-phosphonate in water with the aid of triethylamine,
the corresponding dialkyl polyfluoroalkyl 2-hydroxyphosphonate was formed. The reaction run
smoothly and large amounts of thermal energy were released. The hydroxyphosphonates were
obtained as racemates. Single distillation was sufficient for purification for further steps. The
hydroxyphosphonates were subsequently reacted with O-phenyl chlorothionoformate, yielding
xanthates, which were purified by short flash chromatography.
To finally release the desired polyfluorinated phosphonates we attempted to apply standard
Barton-McCombie reaction protocol with tributyltin as a hydrogen source. However, the
organotin by-products were found to be virtually impossible to remove through extraction and
distillation. Any attempt of chromatography led to a rapid decomposition of the product. Thus,
the organotin hydride was replaced with tris(trimethylsilyl)silane. Most of the by-products were
removed by dissolving the dried post-reaction residue in acetonitrile and washing with hexane
12

Page 13 of 24
New Journal of Chemistry
1
2
3
4
5
View Article Online
DOI: 10.1039/D0NJ00963F
or petroleum ether (pentane gave unsatisfactory results). The crude products were purified by
two subsequent distillations.
6
7
8
2.6.Biological activity
The encouraging ADME results for ZOTs, especially, the BBB permeability and lack of
interaction with P-gp, prompted us to study their biological activity in cells derived from
brain.17, 21 We decided to screen the cytotoxic properties of ZOTs in routinely used glioblastoma
U87-MG cell line. Moreover, close but nonfluorinated structural analogs of the ZOTs, dialkyl
(isothiocyanatomethyl)phosphonates, were demonstrated to have antiproliferative activity
against some cancer cell lines including lung (A549), breast and colon cancers.29 To evaluate
whether ZOTs show tissue specificity, apart from U87-MG cells, we studied A549 lung
carcinoma (solid cancer) cells and HL-60 promyelocytic leukemia (blood cancer) cells. The
normal primary human peripheral blood mononuclear cells (PBMCs) were studied to assess the
safety profile of ZOTs. Standard resazurin-based cytotoxicity assay was used. This assay is
based on the reduction of weakly fluorescent resazurin to highly fluorescent resorufin by
mitochondrial dehydrogenases of viable cells. The intensity of fluorescence corresponds to the
number of live cells in a culture.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
The cytotoxicity of a similar group of compounds has not been evaluated elsewhere. Their
structure containing both fluorine, which usually increases biological activity, and phosphonic
moiety, which mimics carboxyl group, indicated that this group of compounds is a promising
candidate to study anticancer properties. In the attempt to gain the structure-activity relationship
in this group, ZOT3-1-Me, ZOT3-1-Et, ZOT5-1-Me, ZOT5-1-Et, ZOT3-2-Me, ZOT3-3-Me
were tested.
The analysis revealed that the pentafluorinated compounds (ZOT5 group, i.e. ZOT5-1-Me and
ZOT5-1-Me) but not trifluorinated compounds ZOT3 group, i.e. ZOT3-1-Me, ZOT3-1-Et,
ZOT3-2-Me, ZOT3-3-Me) exhibited cytotoxicity toward glioblastoma, lung carcinoma and
promyelocytic leukemia cells (Fig. 5, Tab. 8). Their cytotoxic activity depended on the cell line
and the compounds were the most potent toward U-87 MG cells and the least toward HL-60
cells (Tab. 8). Both compounds demonstrated cytotoxicity toxicity toward PBMCs comparable
to that observed in cancer cell lines tested.
Table 8. Cytotoxicity (IC50 ± SEM (95%CI)) of ZOT5-1-Me and ZOT5-1-Et.
Cell line
Tissue / Disease
Me (µM)
Et (µM)
U-87 MG Brain / Glioblastoma astrocytoma
19.30 ± 0.01
11.74 ± 0.01
(18.43 – 20.15)
(11.13 – 12.37)
A549
Lung / Non-small-cell lung carcinoma
Blood / Promyelocytic leukemia
Blood / -
71.62 ± 0.03
68.66 ± 0.02
(60.98 – 84.12)
(62.86 – 74.99)
HL-60
PBMCs
144.9 ± 0.04
84.81 ± 0.03
(121.0 – 173.5)
(75.42 – 95.37)
66.19 ± 0.03
46.10 ± 0.06
(58.83 – 74.46)
(40.28 – 52.77)
13

New Journal of Chemistry
Page 14 of 24
1
2
3
4
View Article Online
DOI: 10.1039/D0NJ00963F
A
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
B
Figure 5. The cytotoxic properties of ZOT5-1-Me and ZOT5-1-Et toward normal human
peripheral blood mononuclear cells (PBMCs) and human cancer cells: non-small cell lung
cancer cells (A549), promyelocytic leukemia cells (HL-60) and glioblastoma astrocytoma cells
(U-87 MG). (A) Cell viability expressed as a percentage of control cells after 24 h of ZOT5-1
treatment. (B) Changes in U-87 MG cell morphology after 24 h of ZOT5-1 treatment.
14

Page 15 of 24
New Journal of Chemistry
1
2
3
View Article Online
The results indicate that pentafluorinated phosphonates of this class exhibit cytotoxic properties
and thus have therapeutic potential as anticancer agents. Diverse sensitivity towardDcOaI:n10c.e10r39c/eDl0lNsJ00963F
to dimethyl and diethyl derivative suggests the existence of a specific molecular target for this
class of molecules. We could only speculate on its nature because cancer cell lines used in this
study have different genetic constitution and origin (solid vs blood cancer). The discrepancy in
the cytotoxicity between blood cells, for ZOT5-1-Me IC50 values 144.9 µM for HL-60 cells
and 66.19 µM for PBMCs, suggests that the insensitivity of HL-60 cells to ZOT5-1-Me and
ZOT5-1-Et could be associated with their deficiency of p53 protein or their undifferentiated
state. Dysfunction of tumor suppressor protein p53 is a common feature in the majority of
human cancers since it controls apoptosis, cell cycle, and DNArepair. Therefore the mechanism
of their action should further be investigated including the role of p53 and differentiation
pathways. Both ZOT5-1-Me and ZOT5-1-Et showed to be the most effective in U-87 MG
glioblastoma cells.
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Since chemotherapeutics are predominantly administered intravenously, they can evoke a
systemic effect of antitumor therapy. The PBMCs were used in this study for the evaluation of
ZOTs cytotoxicity in primary normal cells to give an indication of possible systemic toxicity.
The studied ZOTs demonstrated the cytotoxic effect towards the PBMCs, while high, was
significantly lower than that observed in U87-MG cells (ZOT5-1-Me: 66.19 µM vs 19.30 µM,
p < 0.001 and ZOT5-1-Et: 46.10 µM vs 11.74 µM, p < 0.001).
Currently, a chemotherapeutic regime for glioblastoma multiforme includes the administration
of alkylating agents: temozolomide or carmustine wafer30. Temozolomide and carmustine show
IC50 = 330 µM and IC50 = 632 µM, respectively in U-87 MG cell line after 24 h31,32 Both ZOT5
molecules showed significantly lower cytotoxic concentration than current chemotherapeutics
in U-87 MG cells after 24 h treatment, IC50 = 19.3 µM and IC50 = 11.74 µM, respectively.
Temozolomide and carmustine both belong to alkylating agents, which generally are highly
effective but show weakly cytotoxic properties33.
3. Future perspectives
ZOTs constitute a new group of low molecular weight compounds combining polyfluoroalkyl
and phosphonic moieties. Their unique chemical structure and simplicity hinder the prediction
of the molecular mechanism of their cytotoxic properties since there is no study on biological
targets of similarly structured compounds.
Although both ZOT5-1-Me and ZOT5-1-Et do not meet the nanomolar rule commonly
considered as necessary for further preclinical studies, they demonstrated promising ADME
predicted pharmacokinetic properties and constitute a new group of compounds that could be
chemically modified in the purpose of increasing their cytotoxic properties. The scope of the
investigated compounds could be extended in the future, especially to include longer, fully and
partially fluorinated chains as well as different substituents at the phosphonic site. Therefore,
we believe that this group of compounds deserves further investigation to define their molecular
targets, safety, metabolism, pharmacokinetics, and pharmacodynamics. The observed toxicity
toward PBMCs indicates that for the further research encapsulation before in vivo
administration would be beneficial to lower the systemic toxicity. Simultaneously, the cytotoxic
properties of ZOTs should be screen on the wide spectrum of cancer cell lines, including NCI-
60 panel and glioblastoma cancer cell lines to select the most sensitive cancer types and thus
reduce the off-target toxicity.
Combining the promising cytotoxic effect against the U-87 MG cell line of ZOT5-1-Me and
ZOT5-1-Et, with the results of the ADME profiling suggesting high bioavailability of these
compounds we strongly suggest further investigations of the compounds of this class.
15

New Journal of Chemistry
Page 16 of 24
1
2
3
4
View Article Online
DOI: 10.1039/D0NJ00963F
4. Conclusions
5
6
7
8
In silico ADME profiling predicted high bioavailability of the simplest dialkyl
polyfluoroalkylphosphonates. This includes high gastrointestinal (GI) absorption and blood-
brain barrier (BBB) permeability supported by non-inhibition of the main xenobiotics
biotransformation executors.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Through carefully designed kinetic experiments we proved that the title compounds could not
be obtained by standard Michaelis-Arbuzov or Michaelis-Becker procedures since the reactions
run several orders of magnitude slower than the side reactions. We further confirmed this by in
silico DFT calculations which agree with the experimental data up to the internal uncertainty
of the theoretical method.
To overcome the problem with the availability of the compounds through an organic synthesis
we developed a three-step protocol. The synthesis path starts from the addition of H-
phosphonates to readily available perfluorinated aldehydes and follows Barton-McCombie
deoxygenation, yielding the final products.
We have shown that the simplest pentafluorinated phosphonates exhibit significant toxicity
towards several cancer cell lines including human non-small cell lung cancer (A549), human
promyelocytic leukemia (HL-60) and human glioblastoma astrocytoma (U-87 MG). The latter
cell line was found to be especially sensitive for ZOT5-1-Et and ZOT5-1-Me. The IC50 values
revealed to be as low as 11.74 µM and 19.30 µM, respectively, which are lower than those of
known chemotherapeutics clinically applied in the treatment of glioblastoma.
5. Acknowledgments
5.1.Financial and infrastructural support
Polish National Science Centre (Narodowe Centrum Nauki) is kindly acknowledged for
providing the financial support for this research under the grant number UMO-
2014/13/N/ST5/01532.
This research was supported in part by PL-Grid Infrastructure.
Dr. Emilia Obijalska (Department of Organic and Applied Chemistry, University of Lodz) is
kindly acknowledged for fruitful scientific discussions and technical help. Dr. Paulina Wach
(Institute of Organic Chemistry, Polish Academy of Science) is acknowledged for technical
discussions.
5.2.Competitive interests
Authors have no competing interests to declare.
6. Materials and methods
6.1.Chemical syntheses
All chemicals were purchased from commercial sources and used without further purification
except for solvents which were routinely distilled prior usage. Tetrahydrofuran (THF) was dried
by distillation from sodium/benzophenone, dichloromethane was dried by distillation from
calcium hydride.
16

Page 17 of 24
New Journal of Chemistry
1
2
3
4
5
View Article Online
DOI: 10.1039/D0NJ00963F
The analytical NMR spectra were recorded on a Bruker AVANCE III 600 spectrometer in
deuterated solvents. The chemical shifts (δ) are reported relative to the pre-assigned solvent
peak position (for 1H and 13C{1H}) or external standard (19F and 31P{1H}).
6
7
8
9
Mass spectrometry analyses were performed as a commercial service at the Institute of Organic
Chemistry, Polish Academy of Science, using Synapt G2-S mass spectrometer (Waters)
equipped with the electrospray ion source and quadrupole-Time-of-flight mass analyzer. The
measurement was performed in positive ion mode with the resolving power of the TOF analyzer
20000 FWHM. The exact mass measurements for all peaks were performed within 3 mDa mass
error.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
6.2.Kinetic experiments
Three solutions of MeI in P(OMe)3 were prepared, containing 1, 2, and 4 mol% of MeI
respectively. A sample of pure phosphite was used as a reference and monitored under the same
conditions as well. Three separate solutions of 10 mol% of 1a, 1b, and 1c respectively in
P(OMe)3 were made.
Each solution was placed in a capillary which was sealed by melting the tip. Each capillary was
placed in an NMR tube containing 0.4 ml of D2O. For each solution, the starting point spectra
31
(1H and P) were registered. All tubes were placed in an oil bath at 50 °C and kept at this
temperature for the duration of the experiment. Before every measurement, the NMR
spectrometer was thermostated at 50 °C using a pure D2O sample to minimize the variations of
the temperature of the reaction mixtures.
The spectra were recorded on Varian GEMINI 2000 spectrometer. The spectra were taken every
2 hours for a period of the first 14 hours and then every 24 hours counting from the starting
point. The longest experiment (using 1a) took 38 days.
6.3.Synthesis
Dimethyl (2,2,2-trifluoro-1-hydroxyethyl)phosphonate (43-Me)
Trifluoroacetaldehyde hydrate (75% solution in water, 5 ml, 46.6 mmol) was placed in a round-
bottom flask equipped with a reflux condenser at ambient temperature. Dimethyl H-
phosphonate (10 ml, 109 mmol, 2.3 eq) was added followed by triethylamine (8 ml, 57.5 mmol,
1.25 eq). Heat release was observed. The mixture was stirred for 1 hr. Volatiles were removed
on a rotary evaporator (10 mmHg, 95 °C). The residue was dissolved in water (50 ml) and
extracted with diethyl ether (5×20 ml). Solvents were removed in vacuo to give the crude
product as white crystalline solid (6.5 g, 34%) which was used without further purification. The
1H, 31P, and 19F NMR spectra were consistent with the literature.
1
Colorless crystals, yield: 66.7%; mp 60-62 °C; H NMR (600 MHz, CDCl3) δ: 4.32 (dq,
J = 12.7, 8.2, 1H), 3.88 (dd, J = 16.9, 10.8 Hz, 6H); 13C NMR (151 MHz, CDCl3) δ: 123.1 (dq,
J = 281.2, 7.4 Hz), 67.6 (dq, J = 163.5, 33.7 Hz), 54.7 (d, J = 6.8 Hz), 54.0 (d, J = 7.1 Hz); 31P
NMR (243 MHz, CDCl3) δ: 16.62 br s, 19F NMR (565 MHz, CDCl3) δ: -73.54 (t, J = 8.1 Hz);
HR-MS (ESI) calcd for C4H8F3O4P [(M + Na)+]: 231.0010; found: 231.0003.
Diethyl (2,2,2-trifluoro-1-hydroxyethyl)phosphonate (43-Et)
Synthesized according to the procedure for 43-Me.
1
Colorless crystals, yield: 92.9%; mp 64-65 °C; H NMR (600 MHz, CDCl3) δ: 4.34-4.23 (m,
5H), 1.39 (t, J = 7.1, 3H), 13C NMR (151 MHz, CDCl3) δ: 123.2 (dq, J = 280.8, 6.6 Hz), 67.7
17

New Journal of Chemistry
Page 18 of 24
1
2
3
4
5
31
View Article Online
DOI: 10.1039/D0NJ00963F
(dq, J = 162.6, 33.3 Hz), 64.2 (dd, J = 102.3, 7.0 Hz), 16.3 (d, J = 5.7 Hz), P NMR (243 MHz,
CDCl3) δ: 14.23 (q, J = 7.7 Hz), 19F NMR (565 MHz, CDCl3) δ: -73.30 (t, J = 8.0 Hz); HR-MS
(ESI) calcd for C6H12F3O4P [(M + Na)+]: 259.0323; found: 259.0318.
6
7
8
9
Dimethyl (2,2,3,3,3-pentafluoro-1-hydroxypropyl)phosphonate (45-Me)
Synthesized according to the procedure for 43-Me.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Colorless oil, yield: 85.0%; bp 91-93 °C at 0.40 Torr; 1H NMR (600 MHz, CDCl3) δ: 4.44 (ddd,
J = 25.3, 13.3, 3.1 Hz, 1H), 3.88 (dd, J = 18.7, 10.9 Hz, 6H); 13C NMR (151 MHz, CDCl3) δ:
118.6 (qddd, J = 287.4, 37.1, 34.5, 14.1 Hz), 112.9 (ddqd, J = 261.6, 255.4, 36.9, 2.0 Hz), 66.3
(ddd, J = 162.9, 32.5, 24.9 Hz), 54.7 (d, J = 6.9 Hz), 54.0 (d, J = 7.2 Hz); 31P NMR (243 MHz,
CDCl3) δ: 17.24; 19F NMR (188 MHz, CDCl3) δ: -82.84 (d, J = 3.9 Hz), -119.32 (d, J = 279.5
Hz), -127.80 (ddd, J = 279.5, 25.3, 2.8 Hz); HR-MS (ESI) calcd for C5H8F5O4P [(M + Na)+]:
280.9978; found: 280.9970.
Diethyl (2,2,3,3,3-pentafluoro-1-hydroxypropyl)phosphonate (45-Et)
Synthesized according to the procedure for 43-Me.
Colorless oil, yield: 96.8%; bp 90-93 °C at 0.45 Torr; 1H NMR (600 MHz, CDCl3) δ: 4.41 (ddd,
J = 25.0, 13.1, 2.9 Hz, 1H), 4.33-4.23 (m, 4H), 1.39 (td, J = 7.1, 2.1 Hz, 6H); 13C NMR (151
MHz, CDCl3) δ: 118.7 (qddd, J = 287.2, 37.2, 34.7, 13.6 Hz), 15.8 (d, J = 5.7 Hz), 113.0 (dqd,
J = 145.9, 36.6, 6.4 Hz), 66.1 (ddd, J = 165.1, 31.8, 24.7 Hz), 64.0 (dd, J = 68.7, 7.1 Hz), 15.8
(d, J = 5.7 Hz); 31P NMR (243 MHz, CDCl3) δ: 14.78 quasi-t; 19F NMR (188 MHz, CDCl3) δ:
-82.77 (d, J = 3.8 Hz), -119.04 (d, J = 279.3 Hz), -127.90 (ddd, J=279.3, 25.1, 2.8 Hz); HR-MS
(ESI) calcd for C7H12F5O4P [(M + Na)+]: 309.0291; found: 309.0286.
O-(1-(dimethoxyphosphoryl)-2,2,2-trifluoroethyl) O-phenyl carbonothioate (53-Me)
Dimethyl (2,2,2‐trifluoro‐1‐hydroxyethyl)phosphonate (43-Me) (1.66 g, 8 mmol) was
dissolved in CH2Cl2 (15 ml) and triethylamine was added (1.4 ml, 9.6 mmol, 1.2 eq). The
solution was cooled in an ice-bath and O-phenylchlorothionoformate (1.2 ml, 8.8 mmol, 1.1 eq)
was added dropwise. After the addition was completed the mixture was allowed to heat up to
the rt for 30 min and stirring was continued for 1.5 hr. Water (15 ml) was added, the mixture
was vigorously stirred for 10 min. The layers were separated, the water layer was extracted with
CH2Cl2 (3×20 ml) and combined organic phases were merged and the volatiles removed on a
vacuum rotary evaporator to yield yellow oil which was purified on flash chromatography to
give the product as a colorless oil (2.2 g, 80%).
Colorless oil, yield: 80.5%; Rf = 0.3 (CH2Cl2); 1H NMR (600 MHz, CDCl3) δ: 7.46-7.43 (m,
2H), 7.34-7.32 (m, 1H), 7.16-7.14 (m, 2H), 6.40 (dq, J = 13.2, 7.9 Hz, 1H), 3.93 (dd, J = 12.9,
11.1 Hz, 6H); 13C NMR (151 MHz, CDCl3) δ: 194.0 (d, J = 6.3 Hz), 153.7, 129.7, 127.1, 121.7
31
(dq, J = 281.3, 5.4 Hz), 121.4, 73.7 (dq, J = 164.9, 34.5 Hz), 54.5 (dd, J = 6.3, 3.6 Hz); P
NMR (243 MHz, CDCl3) δ: 10.72 (q, J = 6.6 Hz); 19F NMR (565 MHz, CDCl3) δ: -69.92 (t, J
= 7.3 Hz); HR-MS (ESI) calcd for C11H12F3O5PS [(M + Na)+]: 366.9993; found: 366.9983.
O-(1-(diethoxyphosphoryl)-2,2,2-trifluoroethyl) O-phenyl carbonothioate (53-Et)
Synthesized according to the procedure for 53-Me.
Colorless oil, yield: 85.4%; Rf = 0.3 (CH2Cl2);1H NMR (600 MHz, CDCl3) δ: 7.44 (dd, J = 8.4,
7.6 Hz, 2H), 7.36 – 7.31 (m, 1H), 7.14 (dd, J = 8.6, 1.0 Hz, 2H), 6.37 (dq, J = 13.2, 7.8, 1H),
4.35 – 4.24 (m, 4H), 1.40 (dt, J = 11.0, 7.1 Hz, 6H); 13C NMR (151 MHz, CDCl3) δ: 194.1 (d,
J = 6.3 Hz), 153.7, 129.7, 127.0, 121.8 (qd, J = 281.4, 5.1 Hz), 121.4, 74.6 (q, J = 34.2 Hz),
73.5 (q, J = 34.2 Hz), 64.4 (dd, J = 6.5, 3.5 Hz), 16.3 (dd, J = 5.6, 4.6 Hz); 31P NMR (243 MHz,
18

Page 19 of 24
New Journal of Chemistry
1
2
3
4
5
19
View Article Online
DOI: 10.1039/D0NJ00963F
CDCl3) δ: 8.03 (q, J = 6.3 Hz); F NMR (565 MHz, CDCl3) δ: -70.36 (t, J = 7.3 Hz); HR-MS
(ESI) calcd for C13H16F3O5PS [(M + Na)+]: 395.0306; found: 395.0300.
6
7
8
9
O-(1-(dimethoxyphosphoryl)-2,2,3,3,3-pentafluoropropyl) O-phenyl carbonothioate (55-
Me)
Synthesized according to the procedure for 53-Me.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Colorless oil, yield: 82.9%; Rf = 0.3 (CH2Cl2); 1H NMR (600 MHz, CDCl3) δ: 7.50 – 7.45 (m,
2H), 7.37 (dd, J = 10.6, 4.3, 1H), 7.18 – 7.12 (m, 2H), 6.56 (ddd, J = 21.6, 13.7, 4.2 Hz, 1H),
3.96 (dd, J = 14.1, 11.1 Hz, 6H); 13C NMR (151 MHz, CDCl3) δ: 193.5 (d, J = 4.6 Hz), 153.7,
129.7, 127.1, 121.3, 118.2 (qddd, J = 287.6, 36.6, 34.4, 10.9 Hz), 111.3 (ddqd, J = 265.0, 256.1,
38.8, 2.0 Hz), 72.8 (dd, J = 36.1, 23.5 Hz), 71.2 (dd, J = 36.1, 23.5 Hz), 54.5 (d, J = 6.1 Hz),
54.4 (d, J = 6.7 Hz); 31P NMR (243 MHz, CDCl3) δ: 11.28 (t, J = 3.3 Hz); 19F NMR (565 MHz,
CDCl3) δ: -83.35, -118.95 (dd, J = 286.3, 4.4 Hz), -122.88 (ddd, J = 286.4, 21.1, 1.9 Hz); HR-
MS (ESI) calcd for C12H12F5O5PS [(M + Na)+]: 416.9961; found: 416.9951.
O-(1-(diethoxyphosphoryl)-2,2,3,3,3-pentafluoropropyl) O-phenyl carbonothioate (55-Et)
Synthesized according to the procedure for 53-Me.
Colorless oil, yield: 84.4%; Rf = 0.3 (CH2Cl2); 1H NMR (600 MHz, CDCl3) δ: 7.50 – 7.44 (m,
2H), 7.36 (t, J = 7.5 Hz, 1H), 7.17 – 7.12 (m, 2H), 6.53 (ddd, J = 21.3, 13.7, 4.3 Hz, 1H), 4.39
– 4.27 (m, 4H), 1.43 (dt, J = 7.0, 7.0 Hz, 6H); 13C NMR (151 MHz, CDCl3) δ: 193.5 (d, J = 4.7
Hz), 153.5, 129.6, 126.9, 121.1, 118.2 (qddd, J = 287.4, 36.6, 34.7, 10.7 Hz), 111.2 (ddqd, J =
264.9, 256.9, 38.6, 1.6 Hz), 72.3 (ddd, J = 163.7, 35.7, 23.3 Hz), 64.3 (t, J = 6.4 Hz), 16.1 (t, J
= 5.6 Hz); 31P NMR (243 MHz, CDCl3) δ: 8.58 (t, J = 3.3 Hz); 19F NMR (565 MHz, CDCl3) δ:
-83.40, -118.99 (dd, J = 286.3, 4.5 Hz), -122.93 (ddd, J = 286.3, 20.9, 2.2 Hz); HR-MS (ESI)
calcd for C14H16F5O5PS [(M + Na)+]: 445.0274; found: 445.0263.
Dimethyl (2,2,2-trifluoroethyl)phosphonate (ZOT3-1-Me)
Xanthate 53-Me (2.52 g, 6.54 mmol) was dissolved in toluene (50 ml) and the solution was
heated to 80 °C. Asolution of tributyltin hydride (2.2 ml, 8 mmol, 1.2 eq) andAIBN (197.0 mg,
1.25 mmol, 0.2 eq) in toluene (10 ml) was prepared at ambient temperature and added in
portions to the stirred hot solution of xanthate over 30 min. The mixture was stirred until no
substrate was observed on a TLC plate (around 10-20 min). The solvent was evaporated in
vacuo. The residue was dissolved in acetonitrile (50 ml) and washed five times with hexanes.
The acetonitrile layer was evaporated in vacuo. The product was distilled on a microdistillation
apparatus twice.
Colorless oil, yield: 87.1%; bp 95-96 °C at 38 Torr; 1H NMR (600 MHz, CDCl3) δ: 3.84 (d, J
= 11.3 Hz, 6H), 2.76 (dq, J = 19.6, 10.7 Hz, 2H); 13C NMR (151 MHz, CDCl3) δ: 123.5 (qd, J
31
= 275.8, 3.3 Hz), 53.1 (d, J = 6.3 Hz), 32.0 (dq, J = 144.4, 31.5 Hz); P NMR (243 MHz,
CDCl3) δ: 18.68 (q, J = 14.0 Hz); 19F NMR (565 MHz, CDCl3) δ: -58.57 (dt, J = 13.8, 10.7 Hz);
HR-MS (ESI) calcd for C4H8F3O3P [(M + H)+]: 193.0241; found: 193.0236.
Diethyl (2,2,2-trifluoroethyl)phosphonate (ZOT3-1-Et)
Synthesized according to the procedure for ZOT3-1-Me.
Colorless oil, yield: 87.6%; bp 103-108 °C at 31 Torr; 1H NMR (600 MHz, CDCl3) δ: 4.20 (dq,
J = 14.5, 7.1 Hz, 4H), 2.60 (dd, J = 38.2, 19.1 Hz, 2H), 1.36 (t, J = 7.1 Hz, 6H); 13C NMR (151
MHz, CDCl3) δ: 123.6 (qd, J = 275.7, 3.3 Hz), 62.8 (d, J = 6.3 Hz), 32.7 (dq, J = 144.0, 31.2
Hz), 16.2 (d, J = 6.2 Hz); 31P NMR (243 MHz, CDCl3) δ: 15.87 (q, J = 13.6 Hz); 19F NMR (565
MHz, CDCl3) δ: -58.53 (dt, J = 13.7, 10.7 Hz); HR-MS (ESI) calcd for C6H12F3O3P [(M +
Na)+]: 243.0374; found: 243.0377.
19

New Journal of Chemistry
Page 20 of 24
1
2
3
4
5
View Article Online
DOI: 10.1039/D0NJ00963F
Dimethyl (2,2,3,3,3-pentafluoropropyl)phosphonate (ZOT5-1-Me)
Synthesized according to the procedure for ZOT3-1-Me.
6
7
8
9
Colorless oil, yield: 65.8%; bp 65-66 °C at 25 Torr; 1H NMR (600 MHz, CDCl3) δ: 3.86 (t, J =
7.6 Hz, 6H), 2.65 (q, J = 19.1 Hz, 2H); 13C NMR (151 MHz, CDCl3) δ: 118.5 (qtd, J = 285.7,
36.1, 13.8 Hz), 113.0 (tqd, J = 254.6, 39.1, 9.4 Hz), 53.0 (d, J = 6.3 Hz), 28.2 (dt, J = 144.3,
24.6 Hz); 31P NMR (243 MHz, CDCl3) δ: 19.29 (t, J = 5.6 Hz); 19F NMR (565 MHz, CDCl3)
δ: -86.33, -112.55 (td, J = 18.6, 5.2 Hz); HR-MS (ESI) calcd for C5H8F5O3P [(M + H)+]:
243.0209; found: 243.0214.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Diethyl (2,2,3,3,3-pentafluoropropyl)phosphonate (ZOT5-1-Et)
Synthesized according to the procedure for ZOT3-1-Me.
Colorless oil, yield: 69.1%; bp 105-110 °C at 30 Torr; 1H NMR (600 MHz, CDCl3) δ: 4.20 (dq,
J = 14.5, 7.1 Hz, 4H), 2.60 (dd, J = 38.2, 19.1 Hz, 2H), 1.36 (t, J = 7.1 Hz, 6H); 13C NMR (151
MHz, CDCl3) δ: 118.5 (dtd, J = 285.6, 36.0, 13.4 Hz), 113.1 (tqd, J = 254.4, 39.1, 9.4 Hz), 62.8
(d, J = 6.4 Hz), 29.0 (dt, J = 143.9, 24.5 Hz), 16.2 (d, J = 6.1 Hz); 31P NMR (243 MHz, CDCl3)
δ: 16.45; 19F NMR (565 MHz, CDCl3) δ: -86.34, -112.63 (td, J = 18.6, 5.3 Hz); HR-MS (ESI)
calcd for C7H12F5O3P [(M + Na)+]: 293.0339; found: 293.0342.
Dimethyl (3,3,3-trifluoropropyl)phosphonate (ZOT3-2-Me)
Sodium hydride (7.14 mmol, 60% suspension) was added to dry DMF (9 ml) and the mixture
was cooled down to 0 °C. Dimethyl H-phosphonate (7.14 mmol, 0.75 ml) was added. The
mixture was allowed to warm to room temperature and was stirred for 1 hr. A solution of 1-
iodo-4,4,4-trifluorobutane (3.59 mmol, 854 mg) in a dry DMF (4 ml) was added. The mixture
was stirred for 24 hr. The DMF was removed on a rotary evaporator. The crude product was
distilled twice.
Colorless oil, yield: 16.1%; bp 96-100 °C at 38 Torr; 1H NMR (600 MHz, CDCl3) δ: 3.78 (d, J
= 10.9 Hz, 6H), 2.41-2.35 (m, 2H), 2.00-1.94 (m, 2H); 13C NMR (151 MHz, CDCl3) δ: 126.4
(qd, J = 276.2, 23.8 Hz), 52.8 (d, J = 6.6 Hz), 27.9 (qd, J = 31.2, 3.8 Hz), 18.1 (dd, J = 147.9,
3.1 Hz); 31P NMR (243 MHz, CDCl3) δ: 30.56; 19F NMR (565 MHz, CDCl3) δ: -67.95 (t, J =
10.2 Hz); HR-MS (ESI) calcd for C5H10F3O3P [(M + H)+]: 207.0398; found: 207.0403.
Dimethyl (4,4,4-trifluorobutyl)phosphonate
Synthesized according to the procedure for ZOT3-2-Me.
Colorless oil, yield: 62.2%; bp 105-110 °C at 38 Torr; 1H NMR (600 MHz, CDCl3) δ: 3.78 (d,
J = 10.8 Hz, 6H), 2.27-2.20 (m, 2H), 1.95-1.82 (m, 4H); 13C NMR (151 MHz, CDCl3) δ: 126.5
(dq, J = 276.4, 1.5 Hz), 33.8 (qd, J = 28.8, 14.7 Hz), 52.1 (d, J = 6.7 Hz), 23.4 (d, J = 143.1
31
19
Hz), 15.3 (dq, J = 6.9, 3.4 Hz); P NMR (243 MHz, CDCl3) δ: 32.67; F NMR (565 MHz,
CDCl3) δ: -66.34 (t, J = 10.7 Hz); HR-MS (ESI) calcd for C6H12F3O3P [(M + Na)+]: 243.0374;
found: 243.0373.
6.4.Theoretical calculations
The calculations were conducted using Gaussian 09 set of codes. The method used was Density
Functional Theory (B3LYP functional with empirical GD3 correction for dispersion
interactions: B3LYP-GD3, together with def2tzvp basis set34 with an ECP pseudopotential35 for
the iodine atom; both adopted from EMSL Basis Set Exchange site36). The chosen basis set
corresponds to the 6-311G(2df,p) Pople basis set.
20

Page 21 of 24
New Journal of Chemistry
1
2
3
4
5
6
7
8
9
View Article Online
DOI: 10.1039/D0NJ00963F
Calculations were made for the condensed phase. The solvent (DMF) was modeled using
CPCM approximation. Specific necessary solvent parameters for correct simulation were
adopted from the literature37 and Gaussian website38 as follows:
EPS=37.21938
RSOLV=2.64737
DENSITY=0.00777737
VMOL=128.737
EPSINF=2.03937
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Thermochemical corrections for reagents were calculated for the temperature of 275.15 K (the
creation of sodium salt) and 298.15 K (ambient reaction temperature).
All TS, as well as structures of reactant and product complexes, were confirmed by vibrational
analysis to be first-order saddle points and minima respectively on the Potential Energy Surface.
The vibrational scaling factor was chosen to be 0.98.
6.5.Cell line and culture
Human glioblastoma astrocytoma (U-87 MG) cell line was purchased from HPA Culture
Collections and supplied by Sigma-Aldrich (Sigma-Aldrich, Poznan, Polska). Human non-
small cell lung cancer (A549) cell line was obtained from American Type Culture Collection
(ATCC, Rockville, MD, USA). Human promyelocytic leukemia (HL-60) was a kind gift from
dr Adrianna Nowak from Lodz University of Technology. Human peripheral blood
mononuclear cells (PBMCs) were isolated using Histopaque 1077 (Sigma-Aldrich, Poznań,
Polska) from leukocyte buffy coat obtained from the Blood Donation Center in Lodz, Poland.
U-87 MG cells were grown in EMEM (Sigma-Aldrich) with 1% non-essential amino acids
(Sigma-Aldrich) and 1 mM sodium pyruvate (Sigma-Aldrich). A549 cells were cultured in
DMEM:F-12 1:1 mixture medium with 15 mM HEPES (Lonza, Basel, Switzerland), HL-60
cells were maintained in IMDM medium with 15 mM HEPES (Lonza) and PBMCs were
cultured in RPMI 1640 medium (Lonza). All media were supplemented with 2 mM L-glutamine
(Lonza), 100 units/ml penicillin (Lonza), 100 μg/ml streptomycin (Lonza) and 10% fetal bovine
serum (FBS, Biowest, Nuaillé, France), except for HL-60 cell medium which contained 15%
FBS. Cells were incubated in standard conditions (37 °C, 5% CO2, relative humidity 100%).
6.6.Cytotoxicity
Compounds cytotoxicity was determined with the resazurin-based in vitro toxicology assay kit
(Sigma-Aldrich, Poznan, Polska). Shortly, 10 000 cancer cells or 50 000 PBMCs in 100 µl
culture medium were seeded on 96-well plate a day before the experiment and then exposed in
triplicate to different concentrations of compounds for 24 h. Tested compounds were prepared
in DMSO and diluted in medium with the final concentration of DMSO lower than 1%. After
the incubation, the resazurin dye solution was added in an amount equal to 10% of the culture
medium volume, and the cells were incubated for another 3 h. Then the fluorescence intensity
was measured with the excitation and emission set at 530/25 nm and 590/35 nm, respectively
using a Synergy HT spectrophotometer (Biotek, Bad Friedrichshall, Germany). The cytotoxic
effect was expressed as the % of negative control cells. All of the results were presented as the
mean ± SEM. Calculations of IC50 and their 95% confidence intervals (95%CI) were conducted
using GraphPad Prism 5 Software (Intuitive Software for Science, San Diego, CA).
7. Literature
21

New Journal of Chemistry
Page 22 of 24
1
2
3
4
5
6
7
8
9
View Article Online
DOI: 10.1039/D0NJ00963F
1 C. Lau, S. Dagnino, K. Kato, X. Ye, A. M. Calafat, G. W. Olsen, J. L.Reiner, B. J. Place,
N. Kudo, Q. Jiang, H. Gao, L. Zhang, B. D. Abbott, H.Viberg, E. Mariussen, D. E. Keil,
C. E. Reed, S. E. Fenton, G. L. Kennedy, J. M. Symons, N. Khalil, M. Lee, K. Steenland,
J. Wambaugh, J. L. Butenhoff, J. V. Rodricks, A. A. Rooney, A. L. Boyles, V. R. Walker,
R. C. Buck, J. C. DeWitt, Toxicological Effects of Perfluoroalkyl and Polyfluoroalkyl
Substances, Humana Press, Cham, 2019.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
2 J. Wang, M. Sánchez-Roselló, J. L. Aceña, C. d. Pozo, A. E. Sorochinsky, S. Fustero V.
A. Soloshonok, H. Liu, Chem. Rev., 2013, 114, 2432-2506.
3 Y. Kobayashi, I. Kumadaki, Acc. Chem. Res., 1978, 11, 197-204.
4 M. Schlosser, Tetrahedron, 1978, 34, 3-17.
5 R. Filler, Y. Kobayashi, Biomedical Aspects of Fluorine Chemistry, Elsevier,
Amsterdam,
1982.
6 J. T. Welch, Tetrahedron, 1987, 43, 3123-3197.
7 G. Resnati, Tetrahedron, 1993, 49, 9385-9445.
8 P. Kafarski, Phosphonates: Their Natural Occurrence and Physiological Role. In
Biological Role of Phosphorus, IntechOpen, 2019.
9 G. Somogyi, P. Buchwald, N. Bodor, Pharmazie, 2002, 57, 135-137.
10 A. Misra, S. Ganesh, A. Shahiwala, S. P. Shah, J. Pharm. Sci., 2003, 6, 252-273.
11 A. A. Kadyrov, E. M. Rokhlin, Bull. Acad. Sci. USSR. Div. Chem. Sci., 1983, 32, 1224-
1230.
12 S. Hyde, J. Veliks, B. Liégault, D. Grassi, M. Taillefer, V. Gouverneur, Angew. Chem.
Int. Ed., 2016, 55, 3785-3789.
13 J. Zhang, X. Huang, R. K. Zhang, F. H. Arnold, J. Am. Chem. Soc., 2019, 141, 9798-
9802.
14 J. Luo, E. J. Han, Q. Shen, M. Huang, Y.G. Huang, H. M. Liu, W. Wang, Q. Y. Chen, Y.
Guo, Org. Process Res. Dev., 2016, 20, 1988-1992.
15 V. S. Ostrovskii, I. P. Beletskaya, I. D. Titanyuk, Org. Lett., 2019, 21, 9080-9083.
16 Bioinformatics, S. I. o. SwissADME. http://www.swissadme.ch/index.php.
17 A. Daina, O. Michielin, V. Zoete, Scientific Reports, 2017, 7, 42717.
18 T. Cheng, Y. Zhao, X. Li, F. Lin, Y. Xu, X. Zhang, Y. Li, R. Wang, L. Lai, J. Chem. Inf.
Model., 2007, 47, 2140-2148.
19 P. Ertl, B. Rohde, P. Selzer, J. Med. Chem., 2000, 43, 3714-3717.
20 J. S. Delaney, J. Chem. Inf. Comput. Sci., 2004, 44, 1000-1005.
21 A. Daina, V. Zoete, ChemMedChem, 2016, 11, 1117-1121.
22 C. A. Lipinski, F. Lombardo, B. W. Dominy, P. J. Feeney, Adv. Drug. Deliv. Rev., 2001,
46, 3-26.
23 A. K. Ghose, V. N. Viswanadhan, J. J. Wendoloski, J. Comb. Chem., 1999, 1, 55-68.
24 D. F. Veber, S. R. Johnson, H. Y. Cheng, B. R. Smith, K. W. Ward, K. D. Kopple, J.
Med. Chem., 2002, 45, 2615-2623.
25 W. J. Egan, K. M., Jr. Merz, J. J. Baldwin, J. Med. Chem., 2000, 43, 3867-3877.
26 I. Muegge, S. L. Heald, D. Brittelli, J. Med. Chem., 2001, 44, 1841-1846.
27 J. B. Baell, G. A. Holloway, J. Med. Chem., 2010, 53, 2719-2740.
28 Wolfram Research, I. Wolfram|Alpha Knowledgebase, Champaign, IL.
https://www.wolframcloud.com/.
29 M. Psurski, K. Błażewska, A. Gajda, T. Gajda, J. Wietrzyk, J. Oleksyszyn, Bioorg. Med.
Chem. Lett., 2011, 21, 4572-4576.
30 L. S. Ashby, K. A. Smith, B. Stea, World J. Surg. Oncol., 2016, 14, 225.
22

Page 23 of 24
New Journal of Chemistry
1
2
3
4
5
6
7
8
9
View Article Online
DOI: 10.1039/D0NJ00963F
31 G. N. Castro, N. Cayado-Gutierrez, F. C. Zoppino, M. A. Fanelli, F. D. Cuello-Carrion,
M. Sottile, S. B. Nadin, D. R. Ciocca, Cell Stress Chaperones, 2015, 20, 253-265.
32 H. F. Chang, W. T. Huang, H. J. Chen, L. L. Yang, Molecules, 2010, 15, 8953-8966.
33 C. C. Wong, K. W. Cheng, B. Rigas, J. Pharmacol. Exp. Ther., 2012, 341, 572-578.
34 F. Weigend, R. Ahlrichs, Phys. Chem. Chem. Phys., 2005, 7, 3297 – 3305.
35 K. A. Peterson, D. Figgen, E. Goll, H. Stoll, M. Dolg, J. Chem. Phys., 2003, 119, 11113-
11123.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
36 Basis Set Exchange (BSE). https://www.basissetexchange.org/.
37 M. Buda, J. Solid State Electrochem., 2013, 17, 3027–3037.
38 http://gaussian.com/scrf/.
23

New Journal of Chemistry
Page 24 of 24
1
2
3
4
View Article Online
DOI: 10.1039/D0NJ00963F
A table of contents entry
5
6
7
8
9
Experimental and computational studies on formation and biological properties of the simplest
polyfluoroalkyl phosphonates
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Piotr M. Zagórski, Paulina Tokarz, Bartłomiej Gostyński, Paweł Tokarz
The simplest polyfluoroalkyl phosphonates exhibit high cell-line specific cytotoxicity