N,N′-Dihydroxy-N,N′-diisopropylhexanediamide, a siderophore analogue, as a possible iron chelating agent for hydroponic conditions: metal equilibrium studies

Article abstract of DOI:10.1007/s13738-017-1057-5

Synthetic iron chelates are the most efficient iron fertilizers. The stability in solution of the iron chelates in the presence of other metal cations that may compete with iron for the chelating agent in nutrient solutions is crucial for its effectiveness in supplying iron to plants. In this work, the chelating properties of N,N-dihydroxy-N,N′-diisopropylhexanediamide (DPH), a biological model of the natural siderophore rhodotorulic acid, were evaluated in order to assess its potentialities for being used, as an iron chelate, in hydroponic cultures. For this purpose, the complexation for the metals [Ca(II), Cu(II), Mg(II), Mn(II) or Zn(II)]–DPH–OH systems has been studied using pH potentiometry and the overall stability constants have been determined for the first time. For all M–DPH–OH systems, MHL and ML species were identified. For [Cu(II), Mn(II) or Zn(II)]–DPH–OH systems, the M₂L₃ species was detected, whereas formation of ML(OH) species was found in the [Ca(II) or Mg(II)]–DPH–OH systems. Finally, the chemical stability of iron chelated with DPH in hydroponic conditions was assessed by computer chemical simulations and compared with the one predicted when iron is chelated with ethylenediamine-N,N′-bis(o-hydroxyphenyl)acetic acid (o,o-EDDHA) in similar conditions.

Full text of DOI:10.1007/s13738-017-1057-5

J IRAN CHEM SOC
DOI 10.1007/s13738-017-1057-5
ORIGINAL PAPER
N,N′‑Dihydroxy‑N,N′‑diisopropylhexanediamide, a siderophore
analogue, as a possible iron chelating agent for hydroponic
conditions: metal equilibrium studies
João G. Martins1 · Carlos M. H. Ferreira1 · Suvendu S. Dey2 · Maria Teresa Barros2 ·
Helena M. V. M. Soares1
Received: 27 May 2016 / Accepted: 7 January 2017
© Iranian Chemical Society 2017
Abstract Synthetic iron chelates are the most efficient
iron fertilizers. The stability in solution of the iron che-
lates in the presence of other metal cations that may com-
pete with iron for the chelating agent in nutrient solutions
is crucial for its effectiveness in supplying iron to plants.
In this work, the chelating properties of N,N-dihydroxy-
N,N′-diisopropylhexanediamide (DPH), a biological model
of the natural siderophore rhodotorulic acid, were evalu-
ated in order to assess its potentialities for being used, as
an iron chelate, in hydroponic cultures. For this purpose, the
complexation for the metals [Ca(II), Cu(II), Mg(II), Mn(II)
or Zn(II)]–DPH–OH systems has been studied using pH
potentiometry and the overall stability constants have been
determined for the first time. For all M–DPH–OH sys-
tems, MHL and ML species were identified. For [Cu(II),
Mn(II) or Zn(II)]–DPH–OH systems, the M2L3 species was
detected, whereas formation of ML(OH) species was found
in the [Ca(II) or Mg(II)]–DPH–OH systems. Finally, the
chemical stability of iron chelated with DPH in hydroponic
conditions was assessed by computer chemical simulations
and compared with the one predicted when iron is chelated
with ethylenediamine-N,N′-bis(o-hydroxyphenyl)acetic acid
(o,o-EDDHA) in similar conditions.
Keywords N,N′-Dihydroxy-N,N′-
diisopropylhexanediamide · Siderophore · Stability
constants · Speciation · Hydroponic · Iron chelates
Introduction
The extreme insolubility of ferric hydroxides, which
reduces the availability of iron (Fe) in soils, is the major
cause of iron chlorosis in plants [1]. Iron chlorosis is mani-
fested by the disappearance of the green colour of the
plants due to a significant decrease of chlorophyll in leaves
[2]. It occurs specially on calcareous soils, where calcium
carbonate buffers soil solution pH in the range of 7.5–8.5
and high bicarbonate concentration is present [3–5]. This
fact affects the development of the plants and decreases the
yield of many crops [2, 4, 6, 7].
Iron, as many other micronutrients, can be provided to
plants by chelated metals. In fact, fertilization with syn-
thetic iron chelates is the most effective agriculture prac-
tice to overcome iron chlorosis [7, 8]. These synthetic
compounds are, however, persistent which might raise
ecological issues. Therefore, the use of more eco-friendly
chelating agents is an alternative to mitigate possible issues
from the synthetic chelates currently used [6].
Siderophores are low molecular weight compounds
(500–1500 Da), possessing organic ligands, which are pro-
duced by some aerobic organisms and plants to acquire
iron under iron-limited conditions [9]. The hydroxamate
group, present in natural compounds, is one of the most
commonly found in the siderophores produced by moulds,
fungi and yeast [10, 11]. Hydroxamic acids take part in
many types of biological systems and are also produced
by several types of plants, mainly to inhibit plant patho-
gens. However, several authors have suggested that these
* Helena M. V. M. Soares
hsoares@fe.up.pt
1
REQUIMTE/LAQV, Department of Chemical Engineering,
Faculty of Engineering, University of Porto, Rua Dr. Roberto
Frias, 4200-465 Porto, Portugal
2
REQUIMTE/LAQV-Department of Chemistry, Faculty
of Science and Technology, Universidade Nova de Lisboa,
Caparica, Portugal
1 3

J IRAN CHEM SOC
Fig. 1 Structure and synthesis of N,N-dihydroxy-N,N′-diisopropylhexanediamide (DPH) using the method described by Smith and Raymond
[15]
compounds are also used by plants for metal ion uptake
from soil [10, 11].
evaluated. Additionally, overall protonation constants were
revisited. Finally, the chemical stability of iron chelated
with DPH in Hoagland nutrient solution was predicted by
chemical simulations to assess the suitability of DPH as a
(green) iron supplier to plants in hydroponic conditions.
This work demonstrated the ability of DPH to maintain
iron in a soluble form up to pH 9.5 under hydroponic con-
ditions. Since DPH is a biological and physicochemical
model of the natural siderophore rhodotorulic acid, this
chelating agent can be an alternative for supplying iron to
plants under hydroponic treatment.
Naturally produced hydroxamates are usually tri-
hydroxamic acids that form very stable 1:1 complexes with
ferric ion. Desferrioxamine B is a microbial hydroxam-
ate siderophore that has been extensively studied and was
found to be an efficient source of iron for many micro-
organisms and for a variety of plants [10, 12].
In contrast to tri-hydroxamic acids, rhodotorulic acid
(RA) is a di-hydroxamic acid, produced by Rhodotorula
pilimane and related yeasts [10]. This siderophore is a
tetradentate ligand, which is unable to satisfy the octahe-
dral coordination of ferric ion through the formation of a
1:1 complex. It had been found to form dimeric complexes
with Fe(III), Al(III) and Cr(III) with the formula M2-(RA)3
[10]. Although the preparation of several N-unsubstituted
di-hydroxamate ligands has been reported [10, 13], they
have not been previously used as siderophore analogues but
rather were prepared for analytical purposes [10].
N,N-Dihydroxy-N,N′-diisopropylhexanediamide (DPH)
(Fig. 1) is a biological and physicochemical model of rho-
dotorulic acid. DPH is constituted by two N-isopropylhy-
droxamic acid groups separated by an aliphatic chain of four
methylene groups. In solution, besides the formation of FeL
(log β = 22.3 [10]), DPH is known to form a very stable
dinuclear species with iron(III), Fe2L3 (log β = 62.2 [10]).
The chelating properties of hydroxamic acids
are extremely important in the biological activity of
hydroxamic acids, and the knowledge of the stability in
solution of new chelating agents is essential to predict their
behaviour as iron fertilizers in hydroponic culture but also
for fertigation and soil applications [6, 11].
The ability of a chelating agent to maintain iron(III) in
solution without being displaced by other cations is essen-
tial for its efficiency to remediate Fe chlorosis. Thus, in
order to evaluate correctly the suitability of DPH to com-
plex iron efficiently under hydroponic conditions, as a new
environmental friendly iron chelate, the interaction of DPH
with the various metal ions typically present in this solu-
tion is of paramount importance. However, this information
is not provided in the literature. Therefore, metal formation
stability constants of DPH with macronutrients [Ca(II) and
Mg(II)] and micronutrients [Cu(II), Mn(II) and Zn(II)],
which are typically present in hydroponic solutions, were
Materials and methods
All reagents used in this work were of analytical grade. All
aqueous solutions were prepared with Milli-Q water with
resistivity of 18.2 MW cm−1. CaCl2, CuCl2, MgCl2, MnCl2
and ZnCl2 standard solutions, as well as KOH and HCl
products, were obtained from Merck.
Synthesis of DPH
All solvents were purified by standard methods and dis-
tilled before use [14]. Elemental analyses were performed
with a Thermo (Massachusetts, EUA) Finnigan-CE Flash
EA 1112 CHNS series analyser. NMR spectra were
recorded either on a Bruker (Massachusetts, EUA) AMX-
400 MHz or on a Bruker (Massachusetts, EUA) Avance
600 MHz apparatus, using as an internal standard, the
residual peak of H2O.
DPH was synthesized according to scheme, described in
Fig. 1, using the method described by Smith and Raymond
[15], as a white powder, with an overall yield of 60% and a
purity of 93%, as its hydrochloride salt. 1H NMR (DMSO-d6):
δ 4.55 (m, 2H), 2.32 (t, J = 6.4 Hz, 4H), 1.48 (t, J = 6.4 Hz,
4H), 1.04 (d, J = 6.6 Hz, 12 H). 13C NMR: δ 177.2, 50.9, 36.9,
29.1, 24.2 Anal. Calcd for C12H26Cl2N204 (%): C, 43.25; H,
7.86; N, 8.41. Found: C, 46.51: H, 7.94: N, 7.86.
Potentiometric studies
The potentiometric titrations were performed with a PC-
controlled system assembled to a Crison (Barcelona, Spain)
1 3

J IRAN CHEM SOC
MicropH 2002 meter, a Crison (Barcelona, Spain) MicroBU
2030 microburette, a Philips (Amsterdam, Netherlands) GAH
110 glass electrode and an Orion 90-02-00 (double junc-
tion) reference electrode (Massachusetts, EUA) with an outer
chamber filled with 0.1 mol L−1 KCl. Automatic acquisition
of data was done using a home-made computer program,
COPOTISY. All measurements were taken on solutions
adjusted to an ionic strength of 0.1 mol L−1 KCl in a Metrohm
(Herisau, Switzerland) jacketed glass vessel equipped with a
magnetic stirrer and thermostated at 25.0 0.1 °C using a
water bath. The measurements were taken under anaerobic
conditions by bubbling purified nitrogen (99.9995% purity)
for ca. 15 min before starting the titration and during the titra-
tion in order to exclude oxygen and carbon dioxide from the
reaction mixture in the titration cell.
where U is the objective function to be minimized, N is the
total number of experimental titration points; np represents
the number of parameters simultaneously optimized, Enobs
is the observed electrode potential at the nth data point, and
ꢀ
ꢁ
′
′
Encalc = Eo + k log H+ , where Eo is the electrode inter-
cept and k the electrode calibration slope.
The Hamilton R-factor, RH, is the statistical parameter
used by ESTA to reflect the improved agreement between
the calculated and the observed data. It is given by:
ꢀ
ꢂ
1/2
U
RH
=
.
(2)
ꢁ
N
n=1 Enobs
The Hamilton R-factor does not have dimensions. The
ESTA software includes a proton, ZH, and a metal, ZM, forma-
The glass electrode calibration, in terms of hydrogen ion
concentration, was accomplished by addition of a standard-
ized solution of potassium hydroxide (KOH) to a standard-
ized solution of hydrochloric acid (HCl) (both solutions
adjusted to the ionic strength of 0.1 mol L−1). From this
potentiometric titration, the values of Eo and response slope
were obtained by fitting a straight line to the experimental
points collected around pH 2 and 11.
tion functions defined by
ꢀ
ꢁ
ꢀ
ꢁ
[HT] − H+ + OH−
[LT]
ZH
=
(3)
(4)
and
ꢀ
ꢄ
ꢂ
ꢃ
ꢁ
N
n
[LT] − L 1 + n=1 βn01 H+
ZM
=
[MT]
The protonation and stability constants of DPH with the
metal ions Ca(II), Cu(II), Mg(II), Mn(II) and Zn(II) were
determined by direct potentiometric titrations. For all sys-
tems, monotonic additions of standardized KOH (titrant)
were made, and the potential was recorded as a function
of the added volume. For the determination of the protona-
tion constants, pH potentiometric titrations were performed
using 2 independent solutions of DPH 2 × 10−3 mol L−1.
For all metal–DPH systems studied, pH potentiometric
titrations were performed using two fixed ligand-to-metal
ion ([LT]:[MT]) molar concentration ratios with different
metal concentrations: for Cu(II)–DPH system and Zn(II)–
DPH, metal concentrations of 1.0 × 10−3 and 2.0 × 10−3
mol L−1 were tested at [LT]:[MT] = 2 and [LT]:[MT] = 4 in
a total of 12 and 11 titrations, respectively; for the Mn(II)–
DPH system, [LT]:[MT] = 2 and [LT]:[MT] = 4 ratios were
also used, but metal concentrations of 1.25 × 10−3 and
2.5 × 10−3 mol L−1 were tested in a total of 8 titrations. For
Ca(II)–DPH and Mg(II)–DPH systems, [LT]:[MT] = 1 (metal
concentrations of 2.0 × 10−3 and 4.0 × 10−3 mol L−1) and
[LT]:[MT] = 2 (only for 4.0 × 10−3 mol L−1 of metal con-
centration) ratios were tested for a total of six titrations.
The simulation and optimization procedures of potentio-
metric data were performed using the equilibrium simula-
tion for titration analysis (ESTA) program [16, 17].
where
ꢀ
ꢁ
ꢀ
ꢁ
[HT] − H+ + OH−
L =
ꢀ
ꢁ
ꢂ
(5)
n
N
n=1 βn01 H+
[HT], [LT] and [MT] represent the total concentration of pro-
ton, ligand and metal, respectively; βn01 stands for the forma-
tion constants of MnL0Hl species. ZH and ZM functions are
calculated for each datum point and are plotted against pH
and −log[L], respectively, to aid in the modelling procedures.
For each M–DPH system, several models were tested.
When the final model was established, as the best one
that fitted the experimental data, the overall stability con-
stants were refined simultaneously by ESTA using the
data obtained from all titrations. During the refinement
operations, the water protonation constant [18] and all
known stability constants for metal hydroxide species [18],
Mx(OH)y, were kept fixed (Table 1).
Computer chemical simulations
Metal chemical speciation calculations between each
chelating agent and metal(s) were performed using
MINEQL+ version 4.5 (Environmental Research Software,
Hallowell, ME, USA) [19], a computer program that gen-
erates chemical equilibrium concentrations of all species
being considered in the model by the program reactions.
Computational simulations were performed in aqueous
medium in the pH range between 0 and 14.
A Gauss–Newton method is the adopted approach by
ESTA to minimize the objective function, U:
N
ꢂ
ꢀ
ꢁ
−1
U = N − np
(Enobs − Encalc 2
)
(1)
n=1
1 3

J IRAN CHEM SOC
Table 1 Protonation constants for water and overall stability con-
stants for M(II) complexes with OH−, at 25.0 °C [18]
Equilibrium
Log β µ (mol L−1
)
Water
H2O ↔ H+ + OH−
−13.78 0.1
6.1 0.1
Copper
Cu2+ + OH− ↔ CuOH+
Cu2+ + 2OH− ↔ Cu(OH)2
Cu2+ + 3OH− ↔ Cu(OH)−3
Cu2+ + 4OH− ↔ Cu(OH)24−
Cu(OH)2(s) ↔ Cu2+ + 2OH−
11.8 0.1
14.5 1.0
15.6 1.0
−18.9 1.0
2.7 0.1
Manganese Mn2+ + OH− ↔ MnOH+
Mn2+ + 4OH− ↔ Mn(OH)24−
7.7 0.0
2Mn2+ + OH− ↔ Mn2OH3+
3.4 0.0
Fig. 2 Representative pH titration curves for DPH and metal
[Ca(II), Cu(II), Mg(II), Mn(II) or Zn(II)]–DPH–OH sys-
2+
2Mn2+ + 3OH− ↔ Mn2OH3
18.1 0.0
tems. Ca(II) and Mg(II): [LT]:[MT]
=
2 and 1, respec-
Mn(OH)2(s) ↔ Mn2+ + 2OH− −12.8 0.0
tively, with [MT] = 4.0 × 10−3 mol L−1; Cu(II) and Zn(II):
Zn2+ + OH− ↔ ZnOH+
4.6 0.1
11.1 0.0
16.3 0.0
14.8 0.0
5.0 0.0
[LT]:[MT]
=
2, [MT]
=
1.0
×
10−3 mol L−1
; Mn(II):
Zinc
[LT]:[MT] = 2, [MT] = 2.5 × 10−3 mol L−1; for DPH alone:
[LT] = 2.0 × 10−3 mol L−1. μ = 0.1 mol L−1 (KCl) and T = 25 °C;
a = moles of base (KOH) added per mole of ligand
Zn2+ + 2OH− ↔ Zn(OH)2
Zn2+ + 3OH− ↔ Zn(OH)−3
Zn2+ + 4OH− ↔ Zn(OH)24−
2Zn2+ + OH− ↔ Zn2OH3+
4Zn2+ + 4OH− ↔ Zn4(OH)44+
Zn(OH)2(s) ↔ Zn2+ + 2OH−
Fe3+ + OH− ↔ FeOH2+
27.9 3.0
−15.52 0.0
11.24 3.0
21.5 0.5
30.2 0.0
34.4 0.0
24.5 0.1
48.9 3.0
−38.6 0.0
0.96 0.1
The comparative analysis of the curves of pH versus a,
obtained with DPH alone and with DPH and each metal
ion studied (Fig. 2), evidences that complexation of DPH
with the various metal ions follows different behaviours
and reveals great differences in the stability of the M–DPH
complexes.
Iron
Fe3+ + 2OH− ↔ Fe(OH)+2
Fe3+ + 3OH− ↔ Fe(OH)3
Fe3+ + 4OH− ↔ Fe(OH)−4
2Fe3+ + 2OH− ↔ Fe2(OH)22−
3Fe3+ + 4OH− ↔ Fe3(OH)45−
Fe(OH)3(s) ↔ Fe3+ + 3OH−
Ca2+ + OH− ↔ CaOH+
For Cu(II)–DPH system, the analysis of Fig. 2 evidences
that between pH 4 and 10, the pH versus a curve shifts
from the DPH titration curve, which suggests the formation
of a strong major species. Furthermore, precipitation was
observed above pH 5, most likely due to insoluble Cu(II)
complexes, as it was described for other di-hydroxamic
acids [20]; due to this fact, refinement operations were only
performed over a narrow pH range (between 3.0 and 4.8),
where no precipitation was experimentally observed.
Initially, two models were assumed, models I [CuL]
and II [CuHL and CuL], which ESTA successfully refined
(Table 3). However, a mathematical difference between
the calculated and experimental complexation curves was
observed for the whole refined pH range in both models
(Fig. 3a). For higher pH values (pL values <12.5), the cal-
culated ZM function was below the experimental points,
which evidences the possibility of other CuxLy species. The
formation of M2L3 species by di-hydroxamic acids with
some metal ions has already been reported [20, 21]; there-
fore, Cu2L3 was included in new, model III, which repro-
duces well the experimental results in the pH range consid-
ered (Fig. 3a; Table 3).
Calcium
Ca(OH)2(s) ↔ Ca2+ + 2OH−
−5.29 0.0
1.87 0.1
Magnesium Mg2+ + OH− ↔ MgOH+
4 Mg2+ + 4OH− ↔ Mg4OH4
4+
16.1 3.0
−9.2 0.0
Mg(OH)2(s) ↔ Mg2+ + 2OH−
Results and discussion
Protonation and metal complexation of DPH
The dissociation constants of DPH and Mx–DPHy–(OH)z
complex formation systems [M = Ca(II), Cu(II), Mg(II),
Mn(II) and Zn(II)] have been studied by glass electrode
potentiometry (GEP). Representative pH titration curves
for DPH and for the different metal–DPH systems are
shown in Fig. 2, where a value is the ratio of moles of base
added per mole of DPH present; at a = 0, the ligand is in
the H2L form.
The refined dissociation constants of DPH are listed in
Table 2. The values determined in the present work are in
perfect agreement with those published by Barclay et al.
[10] (pK1 = 9.10; pK2 = 9.79). Due to this reason, no more
potentiometric determinations were conducted.
To test model III, a species distribution diagram (SDD) was
generated for [LT]:[CuT] = 2, [CuT] = 1.0 × 10−3 mol L−1
(Fig. 4a). Complexation between DPH and Cu(II) starts at pH 1,
as CuHL species, followed by the formation of CuL and Cu2L3
species at pH 2.5 and 4.0, respectively. The distribution of species
1 3

J IRAN CHEM SOC
Table 2 DPH protonation constants and stability constants (as log β) for DPH with Cu(II), Zn(II), Mn(II), Mg(II) and Ca(II), refined simultane-
ously for all titrations in ESTA, determined by GEP in 0.1 M KCl at 25 °C (for simplicity, charges were omitted)
Equilibrium
DPH
Metal ions
Cu(II)
Zn(II)
Mn(II)
Mg(II)
Ca(II)
H + L ↔ HL
9.79 0.03
–
–
–
–
–
–
–
–
–
–
2H + L ↔ H2L
18.89 0.03
M + H + L ↔ MHL
M + L ↔ ML
–
18.51 0.02 14.86 0.01 13.18 0.02 12.11 0.01 12.20 0.01
–
14.67 0.01 7.63 0.01 4.37 0.05 2.64 0.01
44.88 0.04 23.52 0.03 15
1.1 0.2
–
2M + 3L ↔ M2L3
M + L + OH ↔ ML(OH)
–
1
–
–
–
–
–
6.360 0.008 5.802 0.009
Number of titrations/number of independent
solutions
4/2
8/4
8/4
8/4
8/4
7/4
[LT]:[MT]a
–
2:4
2:4
2:4
1:2
1:2
a
For additional information, see “Potentiometric studies” section
Table 3 Overall stability constants (as log10β values) for M–DPH systems determined by GEP for one titration in 0.1 mol L−1 KCl at 25º C;
M = Cu(II) or Zn(II), [MT] = 1.0 × 10−3 mol L−1
[LT]/[MT]
Metal ion
Cu(II)
2
Zn(II)
2
Complexes
Model I
Model II
Model III
Model I
Model II
Model III
M + H + L ↔ MHL
M + L ↔ ML
2M + 3L ↔ M2L3
Hamilton R-factor, RH
Number of points
NI
17.89 0.06
18.34 0.01
13.952 0.004
43.09 0.01
0.003
NI
14.43 0.09
14.68 0.02
7.27 0.02
22.38 0.06
0.064
13.91 0.04
14.03 0.03
7.18 0.06
7.26 0.06
NI
NI
NI
NI
0.047
57
0.033
57
0.20
61
0.17
61
57
61
NI not included
is in agreement with the experimental observations, as solution
remained buffered in the pH range 3−5, as a result of the forma-
tion of CuHL, CuL and Cu2L3 (Figs. 2a, 4a). In our study, precipi-
tation was observed at pH values higher than 5; this fact suggests
the formation of other, but insoluble, polynuclear copper–di-
hydroxamic acid complexes, Cu2L3, already suggested by other
authors for other metals with other di-hydroxamic acids [20].
As a supportive technique to further demonstrate the
complexation properties of the Cu(II)–DPH system, UV–
Vis spectrophotometric assays were carried at three differ-
ent pH values where (1) no metal complexation is predicted
(pH 1), (2) the formation of CuHL is predicted (pH 3), and
(3) the formation of CuL is predicted (pH 3.4) (Fig. 5).
Given the limitations presented by the early precipitation
of Cu(II) complexes, no further points could be tested. The
results show a shift in the wavelength to lower values from
the test solution where no complexation was predicted (pH
0.9) to the solutions where complexation was predicted (pH
3.0 and pH 3.4). This further testifies the existence of com-
plexation of copper with DPH in solution.
For Zn(II)–DPH system, the analysis of the pH versus
a curve (Fig. 2) suggests the formation of at least one spe-
cies, and the gentle sigmoidal shape between pH 6.0 and
9.5 evidences the formation of one major species. Because
precipitation was observed at about pH 10, the refinement
was restricted from pH 4.0 to 9.0. The same models, as for
Cu(II)–DPH system, were tested (ZnL and ZnHL + ZnL)
and refined (Table 3). As at higher pH values (pH > 8), the
ZM experimental data were trending to ZM = 1.5 (Fig. 3b),
and the formation of Zn2L3 species was also considered in
a third model (model III); this model had the best repro-
duction of the experimental data from all models analysed
(Table 3). Therefore, model III was considered the final
model for Zn(II)–DPH system.
To test the final model, SDD was generated for
[LT]:[ZnT] = 2, [ZnT] = 1.0 × 10−3 mol L−1 (Fig. 4b). The
SDD is in agreement with the pH versus a plot (Fig. 2) and
the ZH function (Fig. 3b). ZnHL species starts to be formed
at about pH 5 and is the predominant complex up to pH 7.
ZnL and Zn2L3 species predominate at pH near 8 and above
1 3

J IRAN CHEM SOC
great extent of the denticity of the ligand and of the pH,
being slower at lower pH ranges [22]. The pH versus a
graph evidences the formation of Mn(II) complexes only
at higher pH values (pH > 6; Fig. 2). As the complexation
begins at acidic pH and considering the behaviour of the
previously studied metals, MnHL and MnL species were
included in the first model (model I). Additionally, Mn2L3
species was included in model II. The first model (model
I, Table 4) fitted properly the experimental data below pH
7.8, showing that MnHL and MnL species are indeed pre-
sent, but not at higher pH values. Only model II was able to
fit the whole studied experimental data with a low Hamil-
ton R-factor (Table 4). Although the SDD generated using
[LT]:[MnT] = 2, [MnT] = 2.5 × 10−3 mol L−1 (Fig. 4c)
shows that in the pH range used during the refinement pro-
cedures Mn2L3 species was only present in very low quan-
tities, the perfect adjustment to the experimental data was
only possible with the inclusion of this species. The SDD
shows that complexation starts at pH ca. 5.5, as MnHL,
being this species the major species present up to pH ca.
8.0. Below pH 8.0, only a dim presence of MnL and Mn2L3
species is predicted. Due to the observable oxidation of
Mn(II) to Mn(IV), no further conclusions can be taken for
higher pH values.
In the case of Ca(II)–DPH and Mg(II)–DPH systems, the
analyses of the pH versus a curve (Fig. 3) hardly suggest
complexation between DPH and these metal ions as each
curve almost superposes with the one of the DPH alone. No
complexation seems to occur until neutral to alkaline pH.
Three models were tested for each M–DPH system (Table 4).
Model I, containing ML species, proved to be ineffective to
replicate the experimental data. The observation of ZM func-
tion obtained in the study of the Mg–DPH system (Fig. 3c)
evidences a strong curving at higher pH values (lower pL
values), as well as a poor fitting at lower pH values, suggest-
ing the presence of MgL(OH)x and MgHL species, consid-
ered in model II. However, only model III, with the inclu-
sion of MgL, MgHL and MgL(OH) species, was able to fit
to the experimental data. A similar behaviour was observed
during the refinement procedures for the Ca–DPH system. In
this case, the SDD obtained for the model III (Table 4) with
[LT]:[CaT] = 2, [CaT] = 4.0 × 10−3 mol L−1 (Fig. 4d) shows
that complexation starts at about pH 7, with the formation
of CaHL, and at about pH 9, both CaL and CaLOH species
start being formed simultaneously, although CaL is residual
and most likely transitory. The SDD does not predict precipi-
tation up to pH 14.
Fig. 3 ZM function for a Cu(II)–DPH–OH, b Zn(II)–DPH–OH and
c Mg(II)–DPH–OH systems. The computed ZM functions were cal-
culated from the refinement operations performed for the mod-
els presented in (a, b) Table 3 for Cu(II) and Zn(II) and (c) Table 4
for Mg(II). a, b [LT]: [MT] = 2, [M] = 1.0 × 10−3 mol L−1, c [LT]:
[MgT] = 1, [Mg] = 4.0 × 10−3 mol L−1
9, respectively. Additionally, the SDD predicts precipitation
at pH 10.5, which is in accordance with the experimental
observations.
Experimentally, the study of the Mn(II)–DPH system
was difficult to perform, due to the oxidation of Mn(II) to
Mn(IV), even under oxygen-free atmosphere, which was
maintained during the potentiometric titrations. Changes
in the colour of the solution were observed from colour-
less to brown, at pH ca. 8.5, which limited the refine-
ment exercises up to pH 8.4. The oxidation kinetics of
Mn(II) to Mn(III) with DPH and other similar hydroxam-
ate complexes of Mn(II) is described to be dependent in a
To further validate model III, as the final model
for the Mg–DPH system, a SDD was generated using
[LT]:[MgT] = 1, [MgT] = 4.0 × 10−3 mol L−1 (Fig. 4e).
Analysis of the SDD shows that complexation begins
approximately at pH 7 with the formation of MgHL spe-
cies followed by MgL one. MgLOH species is present in
1 3

J IRAN CHEM SOC
Fig. 4 Species distribution diagrams computed, assuming the
final models, for metal [Ca(II), Cu(II), Mg(II), Mn(II) or Zn(II)]–
DPH–OH systems. a, b Model III, Table 3, [LT]: [MT] = 2,
[MT] = 1.0 × 10−3 mol L−1; c model II, Table 4, [LT]:[MnT] = 2,
[MT] = 2.5 × 10−3 mol L−1, d, e model III, Table 4, [LT]: [MT] = 2
or 1, respectively, with [MT] = 4.0 × 10−3 mol L−1. The dotted lines
indicate the pH value where precipitation was observed. The dashed
line in (c) Mn–DPH–OH system indicates Mn(II) oxidation
greater quantity from pH 10.5 on. This is in perfect agree-
ment with the analyses of ZM plot (Fig. 3c). The SDD also
predicts precipitation at pH around 11, which was observed
experimentally. Finally, both MgL and MgLOH species are
formed in the same range of pH (8.5 < pH < 9.5), which is
in line with the observation of ZM function (Fig. 3b), where
the curves from each species are barely distinguishable
from each other.
Cu2L3, ZnL, Zn2L3, MnL, Mn2L3, MgL and MgLOH spe-
cies, refined in this work, are in the same order of mag-
nitude as those described by Farkas et al. [20] and Szabo
et al. [22] for similar di-hydroxamic acids. However, all
the stability constant values for MHL species, refined in
this work, present about one log unit higher than those
previously reported by the same authors. This may be due
to the higher basicity of the coordinating oxygen atoms
as a consequence of the isopropyl groups attached to the
hydroxamic acid, whereas in the di-hydroxamic acids
studied by Farkas et al. [20], methyl groups were bounded
The final overall stability constants obtained, after refin-
ing simultaneously all titrations of all [LT]:[MT] ratios,
are presented in Table 2. The stability constants for CuL,
1 3

J IRAN CHEM SOC
[CuT] = 3.2 × 10−7 mol L−1, [MgT] 8.0 × 10−4 mol L−1
[CaT] = 1.6 × 10−3 mol L−1, [ZnT] = 1.5 × 10−6 mol L−1
,
,
[MnT] = 4.4 × 10−6 mol L−1). The concentration
of ligand was calculated to be 10% in excess of the
Fe(III) concentration. Thus, a DPH concentration of
1.65 × 10−4 mol L−1 (Fig. 6a, b) was used in order to sat-
isfy the stoichiometry of the dinuclear Fe2L3 species, to
ensure total iron complexation; Fe(OH)3(s) equilibrium
was also introduced as a solubility controller (Table 1).
A comparative analysis of the iron-chelating ability, by
means of theoretical speciation, between EDDHA and
DPH is relevant for this matter as EDDHA is among the
most effective iron chelating agents used as iron fertiliz-
ers in chlorosis treatment and its use as commercial Fe
chelates is regulated under EU Directive No. 2003/2003
[23, 24]. With this purpose, SDD was also generated for
Fig. 5 UV–Vis spectrum registered for three solutions
of Cu(II) and DPH at three different pH; [LT]: [CuT] = 2,
[CuT] = 1.0 × 10−3 mol L−1
ethylenediamine-N,N′-bis(o-hydroxyphenyl)acetic
acid
to the hydroxamic acid. Additionally, the stability con-
stants for the Ca(II)–DPH system reported here (Table 2)
were lower than those reported before. Again, the isopro-
pyl groups may be responsible as Ca(II) has the larger
ionic radius (100 pm) and there may be repulsion between
Ca(II) and the isopropyl groups, whereas for other metals
the effect is less noticeable as the ionic radius is smaller
(<70 pm).
(o,o-EDDHA) in nutrient solution assuming a [LT]:[FeT]
ratio of 1.1 ([EDDHA] = 1.1 × 10−4 mol L−1; Fig. 6c, d).
The stability constants for o,o-EDDHA complexes were
obtained fromYunta et al. (2003) [2]. Because Mn(II)-(o,o-
EDDHA) and Zn(II)-(o,o-EDDHA) chelates do not form
in the presence of Fe(III) [2], the equilibriums between
o,o-EDDHA and Mn(II) or Zn(II) were not considered.
The analysis of SDDs for EDDHA (Fig. 6c, d) reveals
that even though a minor complexation between EDDHA
and Mg(II) is predicted in the pH of interest (between pH
7.5 and 8.5; Fig. 6c), this fact does not disturb the effi-
cacy of EDDHA for iron complexation for the [LT]:[FeT]
ratio evaluated (Fig. 6d). On the other hand, SDDs for
DPH (Fig. 6a, b) evidence no appreciable complexation
between Ca(II) and Mg(II) and DPH and Fe(III)–DPH
species predominate throughout the agronomic pH range.
These results suggest that DPH is more specific for Fe(III),
under these conditions, than EDDHA. SDD also reveals
that iron is chelated to o,o-EDDHA in a slightly wider pH
Stability of Fe–DPH complexes under hydroponic
conditions
The ability of a chelating agent to maintain iron(III)
in solution without being displaced by other cations is
essential for its efficiency to remediate Fe chlorosis. To
evaluate the ability of DPH, as a potential iron supplier
to plants under hydroponic conditions, chemical simula-
tions were performed considering the composition of the
Hoagland nutrient solution ([FeT] = 1.0 × 10−4 mol L−1
,
Table 4 Overall stability constants (as log10β values) for M–DPH systems determined by GEP for one titration in 0.1 mol L−1 KCl at 25 °C;
M = Mn(II), Mg(II) or Ca(II)
[LT]/[MT]
Metal ion
Mn(II)
2
Mg(II)
1
Ca(II)
2
[MT] (mol L−1
)
2.5 × 10−3
Model I
4.0 × 10−3
Model I
4.0 × 10−3
Model I
Complexes
Model II
Model II
Model III
Model II
Model III
M + H + L ↔ MHL
M + L ↔ ML
13.45 0.01 13.51 0.01 NI
12.17 0.02 12.06 0.01 NI
11.29 0.04
11.39 0.02
1.36 0.03
NI
4.42 0.02 3.85 0.07 2.99 0.06 NI
2.54 0.01 2.17 0.15 NI
NI NI NI
6.48 0.01 NI 5.238 0.007 5.269 0.006
2M + 3L ↔ M2L3
NI
16.82 0.02 NI
NI
M + L + OH ↔ ML(OH) NI
Hamilton R-factor
NI
NI
6.52 0.01
0.02
0.013
0.005
30
0.097
183
0.004
183
0.102
138
0.005
138
0.004
138
Number of points
30
183
1 3

J IRAN CHEM SOC
Fig. 6 Species distribution diagram of the a, b DPH and c, d
EDDHA systems, in Hoagland nutrient solution (for detailed
information about the composition of the solution see text). a, b
[L] = 1.65 × 10−4 mol L−1 and c, d [L] = 1.1 × 10−4 mol L−1. % is
the percentage of species present, a, c with total ligand concentration
set at 100% or b, d with total Fe(III) concentration set at 100%; the
vertical line indicates Fe(OH)3 precipitation
range than DPH. Ca(II) and Mg(II) ions are present in the
Hoagland nutrient solution at greater quantity than Fe(III),
and thus, they can compete with Fe(III) for complexation
sites. Because these ions form less stable complexes with
DPH than Fe(III), both metals complex DPH at a higher
pH range (pH > 10), leaving iron(III) totally complexed
with DPH up to pH 9.5. These results suggest that DPH
fulfils the chelating properties to maintain iron in a soluble
form over the agronomic pH range. Since DPH is a biolog-
ical and physicochemical model of the natural siderophore
rhodotorulic acid, it can be an alternative iron chelating
agent for hydroponic treatment.
References
1. J.E. Loper, J.S. Buyer, Mol. Plant-Microbe Interact. 4, 5–13
(1991)
2. F. Yunta, S. García-Marco, J.J. Lucena, J. Agric. Food Chem. 51,
5391–5399 (2003)
3. J.J. Lucena, J. Plant Nutr. 23, 1591–1606 (2000)
4. P. Nadal, L. Hernández-Apaolaza, J.J. Lucena, Plant Soil 325,
65–77 (2009)
5. K. Ylivainio, Environ. Pollut. 158, 3194–3200 (2010)
6. J.J. Lucena, A. Gárate, M. Villén, J. Plant Nutr. Soil Sci. 173,
900–906 (2010)
7. M. Villén, A. García-Arsuaga, J.J. Lucena, J. Agric. Food Chem.
55, 402–407 (2007)
8. D.F. Clemens, B.M. Whitehurst, G.B. Whitehurst, Fertil. Res.
25, 127–131 (1990)
Acknowledgements This work was financed by Fundo Europeu de
Desenvolvimento Regional (FEDER) funds through Programa Opera-
cional Factores de Competitividade—COMPETE—and by National
Funds through FCT—Fundação para a Ciência e a Tecnologia—in the
ambit of the project PTDC-AAC-AMB-111206-2009. Carlos M.H.
Ferreira acknowledges a grant scholarship financed by the same project,
PTDC-AAC-AMB-111206-2009. Additionally, J. G. Martins acknowl-
edges a Ph.D. Grant Fellowship (SFRH/BD/64718/2009) financed by
FCT. To all financing sources, the authors are greatly indebted.
9. R.C. Hider, X. Kong, Nat. Prod. Rep. 27, 637–657 (2010)
10. S.J. Barclay, B.H. Huynh, K.N. Raymond, Inorg. Chem. 23,
2011–2018 (1984)
11. E. Farkas, E. Kozma, M. Petho, K.M. Herlihy, G. Micera, Poly-
hedron 17, 3331–3342 (1998)
12. H. Siebner-Freibach, Y. Hadar, S. Yariv, I. Lapides, Y. Chen, J.
Agric. Food Chem. 54, 1399–1408 (2006)
13. S.J. Barclay, P.E. Riley, K.N. Raymond, Inorg. Chem. 23, 2005–
2010 (1984)
1 3

J IRAN CHEM SOC
14. D.D. Perrin, W.L.F. Armarego, Dawn R. Perrin, Purification of
Laboratory Chemicals, 2nd edn. (Pergamon Press Oxford, New
York, 1980)
19. W.D. Schecher, D.C. McAvoy, Minql+, a Chemical Equilibrium
Modeling System, Environmental Research Software (Hallowell,
ME, 2002)
15. W.L. Smith, K.N. Raymond, J. Am. Chem. Soc. 102, 1252–1255
(1980)
20. E. Farkas, É.A. Enyedy, H. Csóka, Polyhedron 18, 2391–2398
(1999)
16. P.M. May, K. Murray, D.R. Williams, Talanta 32, 483–489
(1985)
21. E. Farkas, D. Bátka, Z. Pataki, P. Buglyó, M.A. Santos, Dalton
Trans. 8, 1248–1253 (2004)
17. P.M. May, K. Murray, D.R. Williams, Talanta 35, 825–830
(1988)
18. A.E. Martell, R.M. Smith, Nist Standard Reference Database 46
Version 8.0, Nist Critically Selected Stability Constants of Metal
Complexes Database, US Department of Commerce, National
Institute of Standards and Technology, Gaithersburg (2004)
22. O. Szabó, E. Farkas, Inorganica Chim. Acta 376, 500–508 (2011)
23. P. Rodríguez-Lucena, L. Hernández-Apaolaza, J.J. Lucena, J.
Plant Nutr. Soil Sci. 173, 120–126 (2010)
24. The European Parliament and the Council of the European Union.
Regulation (EC) No 2003/2003 of the European Parliament and of
the Council of 13 October 2003 relating to fertilisers (2003)
1 3