Chinese Chemical Letters 27 (2016) 1523–1530

Contents lists available at ScienceDirect

Chinese Chemical Letters

journal homepage: www.elsevier.com/locate/cclet

Original article

Acid dissociation constants and cytotoxicity test

of a series of omega-aminoalkyl phosphates
Fu-Hua Sun, Yuan-Zhu Long, Xiao-Yong Zhou, Yi-Rou Jiang, Xing-Yi Xie *
Department of Polymeric Biomaterials and Artificial Organs, College of Polymer Science and Engineering, Sichuan University, Chengdu 610065, China

A R T I C L E I N F O A B S T R A C T

Article history: We synthesised a series of v-aminoalkyl sodium hydrogen phosphates (AAP-n-Na, n = 3, 4, 5, 6,

Received 28 January 2016 purity > 99%), which have potential applications as bioactive cosmetic ingredients and surface modifiers
Received in revised form 11 March 2016 of bone minerals (i.e. hydroxyapatites). Results from Fourier transformed infrared (FTIR), nuclear
Accepted 16 March 2016
magnetic resonance (NMR) and high resolution mass spectroscopy, and elemental analysis all matched
Available online 24 March 2016
their chemical structures. The acid dissociation constants (pKa’s) of each AAP-n (acid form of AAP-n-Na,
n = 2–6) were measured by potentiometric titration, showing a general increasing trend with an increase
Keywords:
in the chain length of AAP-n. However, the pKa3 constant, which corresponds to the deprotonation of the
Aminoalkyl phosphates
Dissociation constant
ammonium group in AAP-n-Na, displayed an unusual decrease when n = even. This odd–even effect can
Cytotoxicity be explained by the pairwise self-association of AAP-n-Na molecules in water where intermolecular
Odd–even effect hydrogen bonding in case of n = even is weaker than that in case of n = odd. All AAP-n-Na at
Hydroxyapatite concentrations up to 0.1% (w/v) were non-toxic to L929 fibroblasts and MG 63 osteoblast-like cells in
terms of cell growth and morphology. These basic data were important for applications of AAP-n and
their salts in biomedical engineering.
ß 2016 Chinese Chemical Society and Institute of Materia Medica, Chinese Academy of Medical Sciences.
Published by Elsevier B.V. All rights reserved.

1. Introduction
Nowadays, v-aminoalkyl dihydrogen phosphates H2N–(CH2)n–
OP(O)(OH)2 (referred to as AAP-n thereafter) and their salts have
found limited biomedical applications despite the fact that they
possess functional amino and phosphate groups present in many
biological molecules, like proteins and nucleic acids. AAP-2, a
phospholipid moiety, has been used to stabilise apatite colloid
with potential for cellular drug delivery [1]. AAP-3 (or its salt) is an
active cosmetic ingredient promoting collagen biosynthesis in the
skin, as shown in a US patent [2]. AAP-6 has been condensed with
biological molecules like biotin [3] and uridine 50 -monophosphate
[4] to form various bioconjugates.
The synthesis of AAP-n involves O-selective phosphorylation of
corresponding amino alcohols. Phosphorus oxychloride (POCl3) is
not suitable because it reacts with both hydroxyl and amino
groups. Only AAP-3 has been synthesised using this phosphory-
lation reagent, which reacts with 3-aminopropanol forming a
cyclic phosphoramidate chloride initially, followed by a selective
* Corresponding author.
E-mail address: xiexingyi@scu.edu.cn (X.-Y. Xie).
hydrolysis of the P(O)–N bond in the six-member ring [2]. A more
universal synthesis has been achieved using phosphoric acid [3–6]
or pyrophosphoric acid [7] as the phosphorylation reagent.
However, a high temperature (140–250 8C), a long reaction time
(18–40 h) and a high vacuum (below 50 mmHg) are usually
required for these methods.
The inconvenient synthesis may account for the limited
availability of AAP-n. To the best of our knowledge, only AAP-2
is commercially available (e.g. from Sigma–Aldrich). We recently
synthesised a series of ammonium salts of AAP-n (n = 3, 4, 5, 6) at
mild temperatures (0–25 8C) using POCl3 as the phosphorylation
reagent [8]. The key is to protect the amino group of each amino
alcohol with a fluorenylmethyloxycarbonyl (Fmoc) group prior to
phosphorylation. We further adopted these ammonium salts as
dispersing agents to synthesise hydroxyapatite hydrocolloids [9],
showing an increase in the aspect ratio of the colloidal particles
with an increase in the carbon number of the dispersant.
In this study, we modified our previous synthesis [8], forming a
monosodium salt of each AAP-n (n = 3–6, referred to as AAP-n-Na
thereafter) which are easy to purify via recrystallisation. This
simple synthesis resulted in highly pure AAP-n-Na (purity over
99%), making it possible to finely characterise their chemical
structures and compositions. Based on this, the pKa constants of

http://dx.doi.org/10.1016/j.cclet.2016.03.029
1001-8417/ß 2016 Chinese Chemical Society and Institute of Materia Medica, Chinese Academy of Medical Sciences. Published by Elsevier B.V. All rights reserved.
1524 F.-H. Sun et al. / Chinese Chemical Letters 27 (2016) 1523–1530
AAP-n series and the cytotoxicity of their sodium salts (AAP-n-Na)
were determined and reported for the first time. We believe these
basic data are important for further research into and applications
of AAP-n and their salts in biomedical engineering, such as the
functionalisation of hydroxyapatite, which is the mineral phase
of bone.
2. Experimental
2.1. Materials
O-[4_TD$IF]Phosphorylethanolamine (i.e. AAP-2, >98%, TCI, Japan) was
used as a control. 3-Amino-1-propanol (>98.5%, J & K Scientific
Ltd., Beijing, China), 4-amino-1-butanol (>98%, Chengdu Best
Reagent Co., Ltd., Chengdu, Sichuan, China), 5-amino-1-pentanol
(>96%, Alfa Aesar, USA) and 6-amino-1-hexanol (>97%, also from
J & K Scientific Ltd.) served as amino alcohols (AC-n, n = 3–6).
Fluorenylmethyloxycarbonyl chloride (Fmoc-Cl, 99%, Asta Tech,
Chengdu, Sichuan, China) and POCl3 (>98%, Kelong Chemical,
Chengdu, Sichuan, China) were used as amino-protecting and
phosporylating agents, respectively.
2.2. Synthesis of AAP-n-Na
As shown in Fig. 1, the whole synthesis involved three steps, i.e.,
protecting the amino group, phosphorylating the hydroxyl group
and removing the protecting group. The first two steps were
described previously [8]. In the third step, each Fmoc-AAP-n
(0.025 mol) was dissolved in 30 mL of N,N-dimethylformamide
(DMF), into which 150 mL of piperidine/DMF (1:4, v/v) was slowly
dripped. After magnetically stirring for 2 h, a resulting white
precipitate (which was a mixture of AAP-n and its piperidine salt
with a molar ratio of 1:1, see Fig. S1 in Supporting information) was
obtained by filtration, and then washed with 30 mL  3 of ethyl
acetate. The washed precipitate was dissolved in 30 mL of water,
and the pH was adjusted to about 11.2 using a solution of 0.5 mol/L
NaOH. The water phase was extracted with 20 mL  5 of chloro-
form. Each extraction proceeded for at least 0.5 h under vigorous
stirring to allow piperidine to enter into the organic phase.
Afterwards, the pH of the water phase was adjusted to 8.68.9
using 0.5 mol/L HCl prior to the addition of 20 mL of ethanol. The
[(Fig._1)TD$IG]
O POCl3
H CH2 O C Cl NH2 CH2 nOH
Fmoc-Cl AC-n
Step 3
O DMF/Piperidine
Fmoc NH CH2 nO P OH
OH Remove Fmoc
NaOH Extraction with CHCl3
Auqeous
pH=11.2 to remove piperidine
O Recrystallised at -18 O C
Ethanol
Crude NH3 CH2 nO P ONa
-18 O C for 15 h
O
Fig. 1. Synthesis and purification of AAP-n-Na (n = 3, 4, 5, 6).
mixture was rested at 18 8C for 15 h to crystallise AAP-n-Na.
Finally, the product was recrystallised the same way, and dried at
60 8C for 10 h.
2.3. Structural characterisation of AAP-n-Na
The Fourier transformed infrared (FTIR) spectra were obtained
on a Nicolet 560 IR spectrometer (Nicolet Instruments, USA) using
KBr disks, with a resolution of 4 cm1[3_TD$IF] between 400 cm1 and
4000 cm1. The 1H, 13C and 31P nuclear magnetic resonance (NMR)
spectra were recorded on a Bruker AV II 400 MHz NMR
spectrometer (Bruker Corp., Switzerland). The high-resolution
mass spectra were collected on a Bruker maXis II time-of-flight
mass spectrometer (Bruker Daltonics, USA).
The C, H and N contents in each AAP-n-Na were analysed on a
Euro EA 3000 elemental analyser (Leeman Labs Inc., USA). The P
and Na contents were measured on a VG PQ ExCell inductive coupled
plasma (ICP) emission spectrometer (TJA Corp., USA), using diluted
sample solutions with known accurate concentrations.
2.4. Potentiometric titration of AAP-n-Na
We performed potentiometric titrations on each AAP-n-Na in
water to determine its purity and the pKa constants of
corresponding AAP-n. Commercial AAP-2 was served as the
control to assess the accuracy of the method. Typically, around
0.2 g AAP-n-Na (accurate mass recorded with an analytical
balance) in 30 mL of water was titrated with 0.05 mol/L HCl and
NaOH standard solutions, respectively. Their accurate concentra-
tions were determined by titration with standard Na2CO3
(99.95%, Tianjin Zhiyuan Chemical Reagent Co., Ltd., Tianjin,
China) or potassium biphthalate (99.95%, Kelong Chemical,
Chengdu, Sichuan, China). All titrations were performed manually
under magnetic stirring at room temperature (25  0.5 8C). About
5 s after each titrant addition, a stable pH value of the analyte solution
was measured with a Sartorius PB-10 pH metre (Sartorius AG,
Germany). The detailed condition of each titration is shown in
Table S1 in Supporting information.
Since the analyte mass (ma) and the titrant concentration (Ct)
are hard to maintain for all titrations (Table S1), it is inconvenient
to compare the titration curves in the form of pH vs. real volume of
Step 1 Step 2
O
Fmoc NH CH2 nOH Fmoc NH CH2 nO P OH
OH
Fmoc-AC-n (yield: 80~85%) Fmoc-AAP-n (yield: 85~90%)
O O
NH3 CH2 nO P OH NH3 CH2 nO P O H2N
O O
O O HCl
NH3 CH2 nO P O NH2 CH2 nO P ONa
pH=8.6-8.9
ONa ONa
O
NH3 CH2 nO P ONa
in water/ethanol O
AAP-n-Na (yield: ~60%)
 F.-H. Sun et al. / Chinese Chemical Letters 27 (2016) 1523–1530
consumed titrant (Vt). Therefore, Vt was normalised to an
equivalent volume (V*) that was defined as:
V  ¼ V t C t =na ¼ V t ðC t M a =ma Þ (1)
where na and Ma stand for moles and molar mass of the analyte,
respectively; the plus and minus sign indicate NaOH and HCl as
the titrant, respectively. Thus V* indicates the degree to which the
analyte has been neutralised by the titrant. The pH values are
plotted against corresponding V* to obtain the titration curve. In
case of 100% purity, the absolute value of V* (jV*j) equals 1 at the
first equivalence point; and the real jV*j at this point in the titration
curve stands for the purity of the analyte. According to definition
(1), the value of jVt/V*j is a constant equalling the theoretical titrant
volume consumed by the analyte at 100% purity in each
dissociation step. These values are listed in Table S1 as well,
ranging from 17.35 to 25.58 mL.
Each AAP-n can be regarded as a triprotic acid with three
dissociation constants (Ka1, Ka2 and Ka3). The pKa constants equal
the corresponding pH values when the concentrations of the acid
and its conjugate base are the same (see Fig. 3A in the Results
section). This is the case at the half equivalence point of each
dissociation step, which can be easily identified in the correspond-
ing titration curve.
2.5. Cytotoxicity test
AAP-2 was transformed into AAP-2-Na by neutralising it with
equimolar NaOH. Each AAP-n-Na (0.1 g, n = 2–6) was dissolved in
8 mL of phosphate buffered saline (PBS, pH 7.4, Fisher Scientific,
Beijing, China). The pH was adjusted to 7.4 using 1 mol/L HCl and
the volume was calibrated to 10 mL. The resulting 1% (w/v)
solution was sterilised by filtration through a 0.22-mm filter
(Minipore, USA) and then diluted to 0.1%, 0.05% and 0.01% with cell
culture medium to obtain AAP-n-Na containing media. According-
ly, the PBS was diluted by 10, 20 or 100 times with the cell culture
medium, serving as negative control media. The cell culture
medium was Dulbecco’s Modified Eagle medium (DMEM, Fisher
Scientific) supplemented with 10% foetal bovine serum (FBS), 0.1%
penicillin and 0.1% streptomycin. The equivolume mixture of 30%
H2O2 and the supplemented DMEM served as a positive control
medium.
L929 mouse fibroblasts (West China Centre of Medical Sciences,
Sichuan University, China) suspended in 200 mL of AAP-n-Na
containing media or control media were seeded at 1000 cells per
well into a 96-well plate (Corning, China). To avoid cross
contamination, the positive control group was arranged in a
separate plate. The medium was refreshed every 2 days. After
incubation at 37 8C with 95% air and 5% CO2 for a predetermined
period (1, 3, 5 and 7 days, n = 6 for each culture condition), cell
morphology was observed using an Olympus IX 71 inverted
microscope (Olympus, Japan) followed by assessing cell prolifera-
tion by the MTT (3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetra-
zolium bromide) assay. In detail, the medium was removed and
20 mL of 5 mg/mL MTT solution (Sigma, USA) was added. After
incubation for additional 4 h in the dark, the supernatant was
discarded and 150 mL of dimethyl sulfoxide was added to dissolve
the purple formazan crystals generated by the cells. Ten min later,
the optical density (OD) of the solution at 570 nm (which is
proportional to the viable cell number) was measured on a
Multiskan MK3 microplate reader (Thermo Scientific, Shanghai,
China).
The relative cell growth rate (RGR) is defined as the percentage
of the OD value of each specimen relative to the OD value of the
corresponding negative control. According to ISO 10993-5, an RGR
of less than 75% indicates some toxicity to cells.
1525
3. Results and discussion
3.1. Chemical structures of AAP-n-Na
As shown in Fig. 2, the FTIR spectra of AAP-n-Na demonstrate
1
absorptions related to phosphate (nP5 5O at 1092 cm , nP–O–C (as) at
982 cm , nP–O–C (s) at 815 cm , dO–P–O (as) at 581 cm1 and dO–P–O
1 1
(s) at 547 cm
1
) [10,11] and alkyl ammonium groups (nN–H, very
broad from 2000 cm1 to 3200 cm1, dN–H (as) at 1635 cm1 and dN–
H (s) at 1587 cm
1
) [12,13]. The free amine (nN–H at 3422 cm1) [12]
and hydroxyl (nO–H at 2571 cm1, from P–OH groups) [10]
absorptions are consistent with the non-inner salt form of AAP-
n-Na. The relative intensity of –CH2– absorptions (nC–H (as) at
2938 cm1, nC–H (s) at 2868 cm1[5_TD$IF] and dC–H at 1484 cm1) increased
with the carbon number in the AAP-n-Na series. The 1H NMR
spectra clearly show every –CH2– signal from each AAP-n-Na with
the same intensity, except that signals e and f from AAP-6-Na
overlap and thus show double intensity. Likewise, every carbon in
AAP-n-Na displayed a separate signal in the corresponding 13C
spectrum. All AAP-n-Na demonstrated a single 31P signal at about
4 ppm because only one type of phosphate group exists in these
molecules. Note that the detailed chemical shifts are summarised
in Table S2 in Supporting information.
The high resolution mass spectra of all AAP-n-Na (Fig. S2 in
Supporting information) mainly showed one peak from corre-
sponding [AAP-n] anions whose absolute m/z value was perfectly
consistent with the molar mass (Table 1). The measured elemental
compositions also highly agree with the theoretical values.
3.2. pKa constants of AAP-n
The pKa constants of the AAP-n series (n = 3–6) were measured
by a simple titration method. To the best of our knowledge, only
the pKa constants of AAP-2 [14,15] and AAP-6 [4] have been
reported previously. Fig. 3A shows the dissociation equations of
AAP-n and the calculation of the dissociation constants. The
titration curves, starting from pure acid form (AAP-2, Fig. 3B) or
monobasic form (AAP-n-Na, Fig. 3C), all display two typical
inflexion points, IP1 and IP2, corresponding to AAP-n and [AAP-n],
respectively. The third inflexion point (IP3) corresponding to [AAP-
n]2 is missing. The inflexion point coordinates and the pKa
constants obtained from the titration curves are listed in
Table 2. The jV*j at the first equivalence point (IP2 for AAP-2 and
IP1 for AAP-n-Na) represents the purity of the titrand which is
above 99% (Table 2). The measured pKa constants of AAP-2 are
consistent with reported values [14], showing the reliability of
the titration method. However, Mohan and Abbott reported
slightly lower values [15]. This might be related to the fact that
KNO3 was added to the titrand solution to adjust the ionic strength
up to 0.2 in that study.
The disappearance of the third inflexion point (IP3, Fig. 3B and
C) in the titration curve of each AAP-n is not surprising. Previous
researchers have also observed the same phenomenon during base
into acid titration of AAP-2 [15] and H3PO4 [16]. Hamann
theoretically analysed the existing conditions for inflexions near
equivalence points in titration curves of polybasic acids [17]. In a
typical base into acid titration of a tribasic acid (pKa1 = 5, pKa2 = 8,
pKa3 = 11), the third inflexion vanishes when the concentration of
the acid or base is less than 0.1 mol/L. In this study, the intervals
between neighbouring pKa constants of AAP-n were about 4–5 pH
units (Table 2), making each dissociation step occur independently,
similar to the typical case proposed by Hamann (where the interval
was 3 pH units [17]). Moreover, the initial concentrations of AAP-n-
Na ranged from 0.031 mol/L to 0.046 mol/L (Table S1) and the
titrant (HCl and NaOH solutions) concentrations were about
0.05 mol/L, i.e. all less than 0.1 mol/L. Therefore, according to the
[(Fig._2)TD$IG]
1526 F.-H. Sun et al. / Chinese Chemical Letters 27 (2016) 1523–1530

Fig. 2. FTIR and NMR spectra of AAP-n-Na. The methylene groups are labelled in (C) for assignments of the 1H and 13
C NMR spectra.

Table 1
Mass spectra and compositional analyses results of AAP-n-Na.

Characterisation n=3 n=4 n=5 n=6


Mass spectra [AAP-n] Calculated (Da) 154.0269 168.0425 182.0582 196.0738
Found (m/z) 154.0265 168.0422 182.0579 196.0739
Combustion elemental analysis (wt%) C Calculated 20.35 25.14 29.28 32.88

Found 19.48 24.95 29.47 33.53

H Calculated 5.12 5.80 6.39 6.90
Found 5.30 5.34 6.41 6.87
N Calculated 7.91 7.33 6.83 6.39
Found 7.49 7.24 6.78 6.59
ICP (wt%) P Calculated 17.49 16.21 15.10 14.13
Found 17.65 16.13 15.01 14.18
Na Calculated 12.98 12.03 11.21 10.49
Found 13.11 11.98 11.16 10.43

theoretical analyses by Hamann, the third inflexions of all AAP-n
should disappear under the conditions used in this study.
3.3. Synthesis art of AAP-n-Na
The FTIR, NMR and high resolution mass spectra, and the
elemental analysis results all highly matched the chemical
structures of AAP-n-Na (Fig. 2 and Table 1), whose purity was
over 99% (Table 2). It can be concluded that the chemical synthesis
in this study was very successful. As a matter of fact, selective
phosphorylation of the hydroxyl groups in a,v-amino alcohols
is the key to synthesising AAP-n. Previous researchers directly
phosphorylated amino alcohols using phosphoric acid or pyropho-
sphoric acid [3–7]. The selection comes from the higher thermal
stability of P(O)–O bond than that of P(O)–N bond at elevated
temperatures (140–250 8C), as suggested in an early patent [5]. We
protected the amino group of a,v-amino alcohols using Fmoc
group before phosphorylation, allowing the synthesis of AAP-n
and their salts at low temperatures (0–25 8C) [8]. Although the
synthesis involves three steps (Fig. 1), the yields of the first two
steps were relatively high (above 80%) and the intermediates
were easily to purify. The first intermediates (Fmoc-AC-n) can be
recrystallised in ethyl acetate/petroleum ether (7:1, v/v) and the
second intermediates (Fmoc-AAP-n) are insoluble in water,
allowing purification by water washing [8].
In the third step, the Fmoc group was removed in piperidine/
DMF, which is a routine method [18]. The released amino groups in
the resultant AAP-n homologues had similar pKb constant (2.80–
2.97, calculated as pKb = 14  pKa3, see Table 2), which approx-
imates the reported pKb value (2.88) of piperidine [19]. This means
[(Fig._3)TD$IG] F.-H. Sun et al. / Chinese Chemical Letters 27 (2016) 1523–1530 1527

A O O O O
Ka1 Ka2 Ka3

=
_ _

=
+ + NH -(CH ) -O-P-OH +
NH3-(CH2)n-O-P-OH 3 2 n NH3-(CH2)n-O-P-O NH2-(CH2)n-O-P-O
_ _

-
_
H+ H+

-
OH O O H+ O
_ _
AAP-n+ AAP-n AAP-n AAP-n 2
_ _
[AAP-n] [ H+ ] [AAP-n ] [ H+ ] [ AAP-n 2 ] [H+]
Ka1 = Ka2 = Ka3 = _
[AAP-n+ ] [AAP-n ] [AAP-n ]

B C
AAP-n AAP-n-Na
12 (n=2) 12 pKa3=11.23
pKa3=10.35 (n=5)

10 10
IP2 -
IP2 AAP-n
8 - 8

dpH/dV*
dpH/dV*
AAP-n pKa2=6.50
pH

pKa2=5.75

pH
6 6

1.002 IP1 0.004

4 IP1 4 AAP-n
pKa1=1.94 AAP-n pKa1=2.17 -0.994
2 -0.002 2

0 0
-1 0 1 2 3 -2 -1 0 1
V* V*

Fig. 3. Dissociation of AAP-n (A) and their representative titration curves (B, C), starting where V* = 0 (the definition of V* is shown in Eq. (1)). The titrand was neutral acid (AAP-
2) (B) or the monosodium salt (AAP-5-Na) (C), both being titrated with HCl (V* < 0) and NaOH (V* > 0) standard solution. Two inflexion points (IP1 and IP2) in each titration
curve are identified at about V* = 0 and about jV*j = 1, representing two equivalence points where the differential curve peaks. The pH value at the half equivalence point of
each dissociation step is the corresponding pKa constant (pKa1, pKa2 or pKa3). Obviously, the pKa1, IP1, pKa2, IP2 and pKa3 points shown in the titration curve are equidistance
horizontally.

Table 2
Titration results of AAP-2 and of AAP-n-Na.

Substance Inflexion point 1 (IP1) Inflexion point 2 (IP2) pKa constants of AAP-n Purity (%)

V* pH V* pH pKa1 pKa2 pKa3

AAP-2 0.000  0.002 3.50  0.04 1.000  0.004 7.94  0.06 2.01  0.06 5.76  0.01 10.43  0.08 100.0  0.4
Ref. [14] – 5.8 10.5
Ref. [15] – 5.52  0.01 10.12  0.01
AAP-3-Na 0.998  0.002 3.92  0.06 0.001  0.001 8.58  0.02 2.16  0.01 6.02  0.03 11.16  0.02 99.8  0.2
AAP-4-Na 0.999  0.003 3.93  0.03 0.002  0.002 8.66  0.04 2.17  0.01 6.43  0.04 11.03  0.03 99.9  0.3
AAP-5-Na 0.997  0.004 4.12  0.03 0.002  0.003 8.81  0.05 2.17  0.01 6.53  0.03 11.20  0.03 99.7  0.4
AAP-6-Na 0.996  0.002 4.15  0.03 0.001  0.002 8.89  0.03 2.16  0.01 6.62  0.02 11.09  0.03 99.6  0.2
Ref. [4] 2.0 6.3 11.0

Three independent titrations were performed for each sample and the titration conditions are shown in Table S1. Note that the theoretical V* values (at 100% purity) for IP1 and
IP2 are 0 and 1 respectively, in case of AAP-2; and the corresponding values are 1 and 0 respectively, in case of AAP-n-Na.

that the primary amino group in AAP-n and the secondary amino obtained by recrystallisation from water/ethanol. Actually, the
group in piperidine possess almost the same potential to be titration results were used to optimise the purification method.
protonated with the dihydrogen phosphate group in AAP- In our previous purification method [8], ammonia solution was
n. Therefore, free AAP-n (in the form of an inner salt) and its used to release piperidine from the piperidine salt of each AAP-
piperidine salt co-precipitated from the DMF solution in a molar n. Ammonia (pKb = 4.74) is a weaker alkaline than piperidine
ratio of about 1:1 (Fig. S1), once the Fmoc group was removed. (pKb = 2.88). This makes ammonia not a good candidate to
The residue of free piperidine was washed away using ethyl deprotonate the piperidine salt. Thus, overdosed ammonia had
acetate. to be used, which was removed from the resultant ammonium salt
For purification, each co-precipitate was dissolved in water, and of AAP-n by freeze-drying. Since no recrystallisation was involved
NaOH solution was added to deprotonate the piperidine salt at pH in that method, the product purity was hard to control. Further
11.2. The free piperidine was extracted with chloroform, leaving to this, it was impossible to obtain pKa constants of AAP-n
the monosodium and disodium salts of AAP-n (Fig. 1) in the homologues by titration of their ammonium salts, since the
aqueous phase. The molar ratio of both salts was about 1:1 because ammonium/ammonia pair can affect the titration curves. Thus, our
the pH value 11.2 approximates the pKa3 constant of AAP-n (Fig. 3C modification to the synthesis of AAP-n salts, compared with the
and Table 2). The pH value was then adjusted to 8.68.9, i.e. close previous method in the literature [8], is fundamental for
to the pH value at the corresponding IP2 point (Fig. 3C and Table 2) applications with AAP-n and their derivatives since we obtained
where AAP-n-Na predominates. Finally, pure AAP-n-Na was a series of pure AAP-n-Na crystals for the first time, making
1528 F.-H. Sun et al. / Chinese Chemical Letters 27 (2016) 1523–1530

possible a systematic investigation of their structures and
properties.
3.4. An odd–even effect on pKa3 constants of AAP-n
An increase in carbon number generally reduced the acidity of
AAP-n because the alkyl group is an electron-donating group. This
is consistent with the general increase trend in both pKa1 and pKa2
values from AAP-2 to AAP-6 (Table 2). The pKa1 constant increased
by about 0.15 from AAP-2 to AAP-3 and reached a plateau at n  3;
the pKa2 constant displayed an obvious increase from AAP-2 to
AAP-4, and then increased very slowly thereafter. Differently, the
pKa3 constant demonstrated a sharp increase from AAP-2 to AAP-3
and then fluctuated at n  3, showing an odd–even effect as a
function of the chain length. In detail, AAP-n with an even carbon
number (e.g. n = 4) possesses a lower pKa3 value than the
neighbouring homologues with odd carbon numbers (n = 3 and
5). Just taking into consideration the AAP-n series with odd carbon
numbers (n = 3, 5) or even carbon numbers (n = 2, 4, 6), a general
increase trend in the pKa3 constant can be observed as well.
Insight into the steric structures of AAP-n-Na with n = odd (3, 5)
and those with n = even (2, 4, 6) can help us to understand the odd–
even effect on the pKa3 value. AAP-3-Na and AAP-4-Na were
selected to illustrate the structural difference between the two
types of AAP-n-Na. Fig. 4A shows their straight chain structures
where all the –CH2– groups are in the trans-conformation to
minimise steric hindrance among them. In this case, the terminal
ionic groups (ammonium cation and phosphate anion) of AAP-3-
Na are in the trans-position relative to the aliphatic chain, while
both terminal ionic groups occupy cis-positions due to one more –
CH2– group in AAP-4-Na.
As there is no externally added salt (e.g. NaCl) to stabilise the
terminal ions in the water where each AAP-n-Na was dissolved,
strong ionic interactions must occur either intramolecularly or
intermolecularly. In the former case, the ionic attraction between
both terminal groups would bend each molecule to form a ring in a
head-to-tail manner, causing server deformations in bond lengths
and bond angles within the molecule and exposing their deformed
hydrophobic –CH2– groups to the water (Fig. 4B). This molecular
conformation therefore is not stable. A more preferable way to
stabilise the terminal ions is to form molecular pairs where every
two molecules are packed in a side-by-side and head-to-tail
manner (Fig. 4C). As a matter of fact, similar pairwise self-
association in water has been proposed for other organic inner
salts, like L-carnitine [20]. In this way, every molecule of AAP-n-Na
can keep its all-trans conformation (Fig. 4A) and its terminal ions
are balanced with counterions from another molecule. The side
hydrophobic interaction from the aliphatic chains and the
hydrogen bonds (see dashed lines, Fig. 4C) between the ammoni-
um and phosphate terminal groups further stabilise the molecular
pairs. The trans-positional terminal groups in AAP-3-Na favours six
hydrogen bonds in each molecular pair as predicted by Materials
Studio software, while only two hydrogen bonds form within each
AAP-4-Na pair due to the cis-positional terminal groups. The extra
hydrogen bonds in AAP-3-Na pairs can provide extra stability to
their ammonium groups. Just taking intermolecular hydrogen
bonding into consideration, the ammonium group in AAP-4-Na is
much easier to deprotonate than that in AAP-3-Na, which favours
a decrease in the pKa3 constant from AAP-3-Na to AAP-4-Na. This
decreasing effect overwhelms the increasing effect due to one
more electron-donating –CH2– group in AAP-4-Na. Overall, the
pKa3 value of AAP-4 is lower than that of AAP-3. Likewise, AAP-6
displays a lower pKa3 constant than AAP-5.
Odd–even effects have previously been observed, mainly in
solid systems where the molecular packing mode depends on the
odd or even number of repeating structural units as well. For
instance, the melting point of straight-chain aliphatic dicarboxylic
acids alternate with the carbon number in the chain [21]. Moreover,
many interfacial properties of organic self-assembled monolayers
(wettability, surface work function, maximum adhesion, tribologi-
cal property, and so on) depend on the odd or even carbon number
in the molecules composing the monolayer, as reviewed in detail
by Tao and Bernasek [22]. The odd–even effect on pKa constant
in water observed in this study is a new phenomenon worth
investigating further.
3.5. Cytotoxicity and biomedical applications of AAP-n-Na
L929 cells generally grew faster in media with AAP-n-Na than in
the negative control media without AAP-n-Na, in particular at day
5 and day 7 (indicated by *, Fig. 5). The variation in tested AAP-n-Na
concentration had no obvious effect on cell growth. The normally

[(Fig._4)TD$IG]

Fig. 4. Possible molecular conformations of AAP-n-Na in water, with AAP-3-Na (top) and AAP-4-Na (bottom) as examples. (A) Straight chain conformations of AAP-3-Na and
AAP-4-Na. (B) Possible ring-like monomeric conformations which are not stable due to server strain in the methylene chain and full exposure of the deformed hydrophobic
chain to water. (C) Molecular pairs stabilised by ionic interactions, hydrogen bonding and hydrophobic interactions. The hydrogen bonds were predicted by Materials Studio
software and are shown as dashed lines. AAP-3-Na obviously possesses stronger intermolecular hydrogen bonding than AAP-4-Na.
[(Fig._5)TD$IG] F.-H. Sun et al. / Chinese Chemical Letters 27 (2016) 1523–1530 1529

1.0 AAP-2-Na
AAP-3-Na
0.8 AAP-4-Na
AAP-5-Na * *

OD Value
0.6 AAP-6-Na
Negative control
Positive control
0.4

0.2

0.0
0.01% 0.05% 0.10% 0.01% 0.05% 0.10%
1d 3d
** *
1.0 * * * **
** *
* **
* ** **
0.8 * *
*
*
OD Value

0.6

0.4

0.2

0.0
0.01% 0.05% 0.10% 0.01% 0.05% 0.10%
5d 7d

Fig. 5. L929 fibroblast growth in media containing AAP-n-Na at various concentrations. Asterisks indicate statistically more cells than the negative control (P < 0.05, Student’s
t test).
[(Fig._6)TD$IG]

Fig. 6. Morphology of L929 cells after 3 days of culture in media containing 0.1% AAP-n-Na.

elongated cell morphology was observed in both negative and
AAP-n-Na media (Fig. 6). On the contrary, spherical cell morphol-
ogy (Fig. 6) and no cell growth (Fig. 5) occurred in the H2O2 solution
(positive control), which was toxic to cells. It can be concluded that
no cytotoxicity is associated with AAP-n-Na at a concentration up
to 0.1%. This was further confirmed by another cytotoxicity test
which showed that 0.5% of AAP-n-Na in cell culture media is toxic
to osteoblast-like MG 63 cells while both 0.01% and 0.1% of AAP-n-
Na are safe to the same cells (Figs. S3 and S4 in Supporting
information). AAP-3 has been reported to be an active cosmetic
ingredient promoting fibroblast proliferation and collagen biosyn-
thesis [2]. Our data further suggest that the AAP-n-Na (n = 2–6)
family might be suitable for cosmetics. The effective concentration
can be as low as 0.01% (Fig. 5).
Previous research has shown that ammonium salts of AAP-n
(n = 2–6) can be used to functionalise hydroxyapatite (bone
mineral) nanoparticles, generating reactive amino groups on them
[9]. These functionalised particles might be grafted with polymers,
1530 F.-H. Sun et al. / Chinese Chemical Letters 27 (2016) 1523–1530
with a final goal to prepare bone-like nanocomposites. AAP-n-Na
can be used in the same way. The measured pKa constants can be
used to calculate the ionic distribution of AAP-n as a function of pH,
which is essential to optimise synthesis art of AAP-n functionalised
hydroxyapatite particles where [AAP-n] ions should predominate
over AAP-n molecules and [AAP-n]2 ions. Taking AAP-5 for
example [AAP-5] ions are over 90% in the pH range of 7.4–10.3
(Fig. 3C). Therefore, AAP-5 functionalised hydroxyapatite should
be synthesised in this pH range.
This study shows that AAP-n-Na homologues are safe to cells at
concentrations lower than 0.1%. When used in vivo, the aminoalkyl
chain bonded hydroxyapatites can slowly release free AAP-n
molecules (mainly in the form of [AAP-n] at biological pH of 7.4,
according to the measured pKa constants, see Fig. 3B and C), which
can be quickly diluted by the body fluid and the real local [AAP-n]
concentration can be much lower than 0.1%. This implies the in vivo
safety of AAP-n and their salts.
4. Conclusions
We synthesised a series of v-aminoalkyl sodium phosphates
(AAP-n-Na, n = 3, 4, 5, 6) with a purity over 99%. Their chemical
structures were confirmed by both spectroscopic and elemental
analyses. The pKa constants of AAP-n were measured by
potentiometric titration, showing a general increase trend with
the increase of chain length in AAP-n. An additional odd–even
effect on the pKa3 constant was observed where AAP-n with
n = even showed a lower pKa3 value than the neighbouring AAP-n
with n = odd. All AAP-n-Na were non-toxic to both L929 fibroblasts
and MG 63 osteoblast-like cells at concentrations up to 0.1% in the
cell media. This fundamental study is essential for the applications
of AAP-n homologues and their derivatives in biomedical
engineering fields.
Acknowledgment
The authors are grateful for financial support from the National
Natural Science Foundation of China (No. 50973069).
Appendix A. Supplementary data
Supplementary material related to this article can be found, in
the online version, at http://dx.doi.org/10.1016/j.cclet.2016.03.029.
References
[1] A. Bouladjine, A. Al-Kattan, P. Dufour, C. Drouet, New advances in nanocrystalline
apatite colloids intended for cellular drug delivery, Langmuir 25 (2009) 12256–
12265.
[2] O.S. Lee, Y.H. Byon, B.S. Lee, et al., Method for preparing 3-aminopropane phos-
phoric acid, US 5723645, 1998.
[3] L.D. Li, W. Tang, W.J. Liu, Z.B. Zhao, Synthesis and activity of the photoaffinity-
labeled functional probe based on prenyl side-chain, Chin. J. Org. Chem. 28 (2008)
489–493.
[4] R. Barker, K.W. Olsen, J.H. Shaper, R.L. Hill, Agarose derivatives of uridine diphos-
phate and N-acetylglucosamine for the purification of a galactosyltransferase,
J. Biol. Chem. 247 (1972) 7135–7147.
[5] F.V. Emiliano, F. Giorgio, Process for preparing ethanol-2-aminophosphoric acid,
US 2730542, 1956.
[6] P.V. Laakso, Method of making 2-aminoethyl phosphate, US 3644603, 1968.
[7] C. Emile, R. Joseph, Improved process for phosphorylating amino-alcohols, US
3098072, 1963.
[8] W.B. Kong, X.Y. Zhou, Y. Yang, X.Y. Xie, A facile synthesis of v-aminoalkyl
ammonium hydrogen phosphates, Chin. Chem. Lett. 23 (2012) 923–926.
[9] Y.R. Jiang, F.H. Sun, X.Y. Zhou, W.B. Kong, X.Y. Xie, Water dispersible hydroxyapa-
tite nanoparticles functionalized by a family of aminoalkyl phosphates, Chin.
Chem. Lett. 26 (2015) 1121–1128.
[10] I. Maege, E. Jaehne, A. Henke, et al., Self-assembling adhesion promoters for
corrosion resistant metal polymer interfaces, Prog. Org. Coat. 34 (1998) 1–12.
[11] D.M. Suflet, G.C. Chitanu, V.I. Popa, Phosphorylation of polysaccharides: new
results on synthesis and characterisation of phosphorylated cellulose, React.
Funct. Polym. 66 (2006) 1240–1249.
[12] K.P. Battjes, A.M. Barolo, P. Dreyfuss, New evidence related to reactions of
aminated silane coupling agents with carbon dioxide, J. Adhes. Sci. Technol. 5
(1991) 785–799.
[13] J.J. Fripiat, M. Pennequin, G. Poncelet, P. Cloos, Influence of the van Der Waals force
on the infrared spectra of short aliphatic alkylammonium cations held on
montmorillonite, Clay Miner. 8 (1969) 119–134.
[14] J. Christoffersen, M.R. Christoffersen, Kinetics of dissolution of calcium hydroxy-
apatite: IV. The effect of some biologically important inhibitors, J. Cryst. Growth
53 (1981) 42–54.
[15] M.S. Mohan, E.H. Abbott, Metal complexes of amino acid phosphate esters, Inorg.
Chem. 17 (1978) 2203–2207.
[16] D.W. Barnum, Predicting acid–base titration curves without calculations, J. Chem.
Educ. 76 (1999) 938–942.
[17] S.D. Hamann, The theory of titration of polybasic acids. II. Conditions for
the existence of inflexions near equivalence points, Aust. J. Chem. 23 (1970)
1749–1765.
[18] J.E. Sheppeck II, H. Kar, H. Hong, A convenient and scaleable procedure for
removing the Fmoc group in solution, Tetrahedron Lett. 41 (2000) 5329–5333.
[19] D.D. Perrin, Dissociation Constants of Organic Bases in Aqueous Solution, Butter-
worths, London, 1965p. 967.
[20] H.Q. Wang, W.N. Cheng, L.Y. Zhu, X.G. Hu, Enthalpic pairwise self-association of
L-carnitine in aqueous solutions of some alkali halides at T = 298.15 K, J. Chem.
Thermodyn. 95 (2016) 111–115.
[21] H.D. Burrows, Studying odd–even effects and solubility behavior using a,
v-dicarboxylic acids, J. Chem. Educ. 69 (1992) 69–73.
[22] F. Tao, S.L. Bernasek, Understanding odd–even effects in organic self-assembled
monolayers, Chem. Rev. 107 (2007) 1408–1453.