Photochemical reactivity of the iron(III) complex of a mixed-donor, α-hydroxy acid-containing chelate and its biological relevance to photoactive marine siderophores

Article abstract of DOI:10.1021/ic500635q

The trimeric clusters [Fe(III)₃(X-Sal-AHA) ₃(μ₃-OCH₃)]^-, where X-Sal-AHA is a tetradentate chelate incorporating an α-hydroxy acid moiety (AHA) and a salicylidene moiety (X-Sal with X being 5-NO₂, 3,5-diCl, all-H, 3-OCH₃, or 3,5-di-t-Bu substituents on the phenolate ring), undergo a photochemical reaction resulting in reduction of two Fe(III) to Fe(II) for each AHA group that is oxidatively cleaved. However, photolysis of structurally analogous mixed Fe/Ga clusters demonstrate that a similar photolysis reaction will occur with only a single Fe(III) in the cluster. Quantum yields of iron reduction for the series of [Fe(III)₃(X-Sal-AHA)₃(μ ₃-OCH₃)]^- complexes measured by monitoring Fe(II) production are twice those for ligand oxidation, measured by loss of the CD signal for the complex due to cleavage of the chiral AHA group. These moderate quantum yields, around 1-2% in the UVA and UVB range, are higher for complexes with electron-withdrawing X groups than for electron-donating X groups. The observed final photolysis product of the chelate is different if irradiation is done in the air than if it is done under Ar. The first observed photochemical product is the aldehyde resulting from decarboxylation of the AHA. This is the final product under anaerobic conditions. In air, this is followed by an Fe- and O₂-dependent reaction oxidizing the aldehyde to the corresponding carboxylate, then a second Fe- and light-dependent decarboxylation reaction giving a product that is two carbons smaller than the initial ligand. These reactivity studies have important biological implications for the photoactive marine siderophores. They suggest that different types of photochemical products for different siderophore structure types do not result from different initial photochemical steps, but rather from different susceptibility of the initial photochemical product to air oxidation.

Full text of DOI:10.1021/ic500635q

Article
pubs.acs.org/IC
Terms of Use
Photochemical Reactivity of the Iron(III) Complex of a Mixed-Donor,
α‑Hydroxy Acid-Containing Chelate and Its Biological Relevance to
Photoactive Marine Siderophores
Jennifer E. Grabo, Mark A. Chrisman, Lindsay M. Webb, and Michael J. Baldwin*
Department of Chemistry, University of Cincinnati, P.O. Box 210172, Cincinnati, Ohio 45221-0172, United States
S
* Supporting Information
ABSTRACT: The trimeric clusters [Fe(III)3(X-Sal-
AHA)3(μ3-OCH3)]−, where X-Sal-AHA is a tetradentate
chelate incorporating an α-hydroxy acid moiety (AHA) and
a salicylidene moiety (X-Sal with X being 5-NO2, 3,5-diCl, all-
H, 3-OCH3, or 3,5-di-t-Bu substituents on the phenolate ring),
undergo a photochemical reaction resulting in reduction of
two Fe(III) to Fe(II) for each AHA group that is oxidatively
cleaved. However, photolysis of structurally analogous mixed
Fe/Ga clusters demonstrate that a similar photolysis reaction
will occur with only a single Fe(III) in the cluster. Quantum yields of iron reduction for the series of [Fe(III)3(X-Sal-AHA)3(μ3-
OCH3)]− complexes measured by monitoring Fe(II) production are twice those for ligand oxidation, measured by loss of the CD
signal for the complex due to cleavage of the chiral AHA group. These moderate quantum yields, around 1−2% in the UVA and
UVB range, are higher for complexes with electron-withdrawing X groups than for electron-donating X groups. The observed
final photolysis product of the chelate is different if irradiation is done in the air than if it is done under Ar. The first observed
photochemical product is the aldehyde resulting from decarboxylation of the AHA. This is the final product under anaerobic
conditions. In air, this is followed by an Fe- and O2-dependent reaction oxidizing the aldehyde to the corresponding carboxylate,
then a second Fe- and light-dependent decarboxylation reaction giving a product that is two carbons smaller than the initial
ligand. These reactivity studies have important biological implications for the photoactive marine siderophores. They suggest that
different types of photochemical products for different siderophore structure types do not result from different initial
photochemical steps, but rather from different susceptibility of the initial photochemical product to air oxidation.
and release the metal upon irradiation.10 These chelates are not
as structurally complex as the natural product marine
INTRODUCTION
■
In environments with limited Fe availability, including waters
near the ocean surface, Fe acquisition can be growth limiting.1,2
siderophores. However, they have more structural variability
and synthetic tunability of their transition metal coordination
In order to sequester sufficient concentrations of bioavailable
iron to sustain life, bacteria produce relatively low molecular
chemistry than other small natural AHAs, such as citrate and
tartrate, which incorporate only other carboxylates among their
weight biomolecules called siderophores that typically bind
Fe(III) with high affinity.3 The ferric stability constants (KML
)
metal-binding functional groups. Among these new AHA-
containing chelates is a series of tetradentate ligands that
incorporate a salicylidene moiety in addition to the AHA.
These chelates have been designated X-Sal-AHA, where X is 5-
NO2, 3,5-diCl, all-H, 3-OCH3, or 3,5-di-t-Bu as substituents on
the phenolate ring. Each of these derivatives produces an
Fe(III) cluster with the same trinuclear structure displayed by
the crystal structure of [Fe3(3,5-diCl-Sal-AHA)3(μ3-OCH3)]−
(Scheme 1).10 These clusters have especially high Fe(III)
stability constants, but release Fe(II) to bidentate receptor
ligands such as phenanthroline upon irradiation.
The AHA-containing chelates provide an alternative
approach to the photoactive chelates referred to as
photocages.11 These photocages frequently take advantage of
the photoactivity of a nitrophenyl group appended to the
for siderophores produced by bacteria in these marine
environments range from as high as log KML = 51 for
alterobactin,4 through marinobactins and aquachelins with log
K
ML’s in the mid-30s,5 to as low as log KML = 24 for
vibrioferrin.6 Some of these marine siderophores undergo
photochemical reactions that result in reduction of the Fe(III)
to Fe(II) and oxidative cleavage of the siderophore. Butler and
co-workers have shown through studies of numerous marine
siderophores that those that are photochemically active include
an α-hydroxy carboxylic acid (AHA) or closely related moiety
among their Fe-coordinating functional groups.3,7 The AHA
moiety is reported to facilitate the photochemical reactivity of
the Fe-bound siderophore through a near-UV ligand-to-metal
charge transfer transition.8,9
Inspired by the photoactive marine siderophores, we have
developed new AHA-containing chelates that are designed to
tightly bind Fe(III) or other transition metals and to reduce
Received: March 19, 2014
Published: May 19, 2014
© 2014 American Chemical Society
5781
dx.doi.org/10.1021/ic500635q | Inorg. Chem. 2014, 53, 5781−5787

Inorganic Chemistry
Article
rophore. Petrobactin8 and aerobactin17 are hydrophilic
examples of the citrate-based siderophores that result in this
kind of product upon irradiation, and the ochrobactins18 and
synechobactins19 are examples of amphiphilic siderophores
within this class. Irradiation of the β-hydroxyaspartate-based
siderophores results in more extensive cleavage. Amphiphilic
examples such as the aquachelins9 lose the fatty acid group
upon irradiation. The alterobactins4 are examples of hydro-
philic, β-hydroxyaspartate-based siderophores. The Fe(III)
stability constants of this class of siderophores tend to decrease
more upon photolysis than do those of the citrate-based
siderophores.3
Scheme 1. Line Drawing of the General Structure of the
10
a
Trinuclear Clusters [Fe3(X-Sal-AHA)3(μ3-OCH3)]−
Why marine bacteria have evolved to produce siderophores
with these different structural properties remains an important
question. Whether the fatty acid moiety is lost and whether
there is a significant decrease in the strength of iron binding
upon irradiation are certain to have significant functional
consequences for the siderophores. A better understanding of
the chemistry that leads to those differences will aid in
answering that question. The results of studies presented here
on the photochemistry of [Fe(III)3(3,5-diCl-Sal-
AHA)3(OCH3)]− will contribute to interpreting the photo-
chemistry of the marine siderophores.
a
The phenolate ring substituents, X, are not included.
chelate. Irradiation either reduces the metal binding ability of
one or more coordinating groups12 or fragments the
chelate.13−15 This approach was initially used with Ca(II) to
examine its biological roles.14,16 More recently it has been
applied to other metals, including the biologically important d-
block metals Cu(II),13 Zn(II),15 and Fe(III).12 In contrast to
the AHA-containing chelates, the photocages are photochemi-
cally sensitive in the absence of bound metal. In some cases, the
quantum yield decreases upon metal binding.13 The stability
constants for metal binding by the photocages also tend to be
rather moderate. The photocage FerriCast, for example, is
highly selective for Fe(III) binding in nonaqueous solvents, but
decomplexation occurs in solvents with oxygen donors such as
water or alcohols.12
Due to the importance of the irradiation products in a
number of potential applications that would employ tight metal
binding with light-triggered metal release, we set out to
thoroughly understand the light-initiated reactivity of the
Fe(III) complexes of the X-Sal-AHA chelates. These chelates
undergo photochemical reactions only when an appropriate
metal, such as Fe(III), is bound. This differs from the
photocages described above, but is similar to the AHA-
containing marine siderophores. The presence of oxygen
significantly affects the observed products after irradiation.
Careful study of the step-by-step reactions that follow
irradiation of the structurally characterized [Fe(III)3(3,5-diCl-
Sal-AHA)3(OCH3)]− cluster was therefore undertaken under
both aerobic and anaerobic conditions. Irradiation of the
Ga(III) and mixed Fe(III)/Ga(III) clusters was also examined
for comparison.
The results of these studies may have significant implications
for interpreting the observed photochemical products of the
natural marine siderophores. Different marine siderophores
produce different kinds of photoproducts, which appear to
correlate with specific structural features.3 Many of the
photoactive marine siderophores incorporate the α-hydroxy
acid group through either a citrate-like moiety in the chelate
backbone or a β-hydroxyaspartate group. Additionally, some
marine siderophores make up suites of amphiphiles that vary
based only on the characteristics of an appended fatty acid
group, while others lack the fatty acid and are better described
as hydrophilic. Both the citrate-based class of marine
siderophores and the β-hydroxyaspartate-based class include
examples of both amphiphilic and hydrophilic members. The
citrate-based siderophores undergo a simple decarboxylation
upon irradiation of the Fe(III) complex. Tautomerization of the
ketone product to the enolate form results in little decrease in
the Fe(III) stability constant compared to the intact side-
EXPERIMENTAL SECTION
■
General Considerations and Synthesis. All solvents and
chemicals were used as received from Fisher/Acros unless otherwise
noted. Anaerobic samples were prepared in an MBraun LabStar
glovebox with an Ar atmosphere and [O2] < 0.1 ppm. The X-Sal-AHA
chelates and their Fe complexes were prepared as described
previously.10 Additional purification to remove salts and any other
impurities was accomplished by suspending the complex in a mixture
of hexane and dibromoethane that matched the density of the
complex, so that any impurities either sank or floated to the surface of
the solution. Solutions of mixed Fe−Ga complexes were prepared by
combining a methanolic solution containing Fe(NO3)3·9H2O and
Ga(NO3)3 with another methanolic solution containing one equivalent
of 3,5-diCl-Sal-AHA and two equivalents of KHCO3 per equivalent of
total metal and stirring overnight. N-(3,5-Dichlorosalicylidene-β-
alanine) (L′) was prepared by the Schiff base condensation of 3,5-
dichlorosalicylaldehyde with β-alanine. Synthetic details for prepara-
tions of L′ and its Fe complexes and their characterization and of the
mixed metal clusters of L are available in the Supporting Information.
Instrumental Methods. UV/visible absorption spectra were
obtained using a Spectral Instruments, Inc. model 420 spectropho-
tometer, using either a fiber optic dip probe with a 1 cm path length or
a cuvette accessory with a 1 cm quartz cuvette. For anaerobic
measurements, samples were loaded into an anaerobic cuvette under
an Ar atmosphere in the glovebox. Circular dichroism spectra were
obtained on a Jasco J-715 spectropolarimeter. The CD intensity was
converted from mdeg to molar absorptivity difference (Δε) using Δε
= θ/32980[M] where θ is the ellipticity in mdeg and [M] is the molar
concentration of the analyte. Electrospray ionization mass spectrom-
etry (ESI-MS) data were collected at the University of Cincinnati Mass
Spectrometry Facility on a Micromass Q-TOF-2 spectrometer.
Photochemical Experiments. Irradiation of samples was
accomplished using a Luzchem LZC 4 V photoreactor with sets of
16 lamps provided by Luzchem Research, Inc. (Ottowa, Ontario,
Canada) centered at 300 nm (UVB), 350 nm (UVA), or 420 nm
(LZC-420). Samples were in 2.5 dram type I, class B borosilicate glass
vials on a rotating, merry-go-round-type sample holder. Lamp intensity
was calibrated using 0.6 mM solutions of K3[Fe(oxalate)3], prepared as
described in the literature, as the actinometer.20 The intensity was
calculated using the following formula: I = [Fe2+] × NA/(Φλ × Δt)
where NA is Avogadro’s number, Φλ is the quantum yield of
K3[Fe(oxalate)3] in the wavelength range of interest, and Δt is time at
1, 3, 7, 15, and 30 s. [Fe2+] in the irradiated sample was found by
5782
dx.doi.org/10.1021/ic500635q | Inorg. Chem. 2014, 53, 5781−5787

Inorganic Chemistry
Article
adding 2.0 mL of 0.1% o-phenanthroline solution and 2.0 mL of
sodium acetate buffer to 4.0 mL of the actinometer solution, then
diluting the solution to 20 mL and measuring the absorbance of
[Fe(phen)3]+2 at 510 nm (ε = 11 100 M−1 cm−1). Quantum yields
were determined by monitoring either Fe reduction or ligand cleavage
at various time points by which photolysis was no more than one-third
complete. The reported values are the average of two separate time
points within the linear part of the time course from each of three
independent trials. The reported error values are the standard
deviation for these six measurements. The concentration of the Fe
complex was 0.3 mM during irradiation with absorbance across the
wavelength range of interest greater than 1, so that >90% of photons
impinging on the sample were absorbed. Iron reduction was
the presence of BPDS, >90% of the Fe is detected as Fe(II)
based on the intensity of the 535 nm feature in the UV/visible
absorption spectrum. After the same amount of time, some CD
intensity remains, as shown in Figure 1. This CD signal is very
4−
monitored by determining the concentration of Fe(II)(BPDS)3 in
solutions that were diluted to 0.04 mM in the presence of excess BPDS
4−
(bathophenanthroline disulfonate). The Fe(II)(BPDS)3 complex
absorbs intensely with a maximum at 535 nm, while there is little
absorbance in the visible range for Fe(III) solutions with BPDS. The
reported values are from experiments done anaerobically to prevent
any air oxidation of the reduced iron. The molar absorptivity of
Fe(II)(BPDS)34− has been reported over a range varying by more than
10%,9,21 so the value used for the calculations reported here was
determined independently in this lab as ε = 20 500 M−1 cm−1 in
methanol, which is in the middle of the range of literature values. The
Figure 1. Circular dichroism spectra of 0.0375 mM methanolic
solutions of Na[Fe3(3,5-diCl-Sal-AHA)3(OCH3)] before irradiation
(thick black solid line), after irradiation under Ar (blue dotted line),
after irradiation under Ar, and then on exposure to air (black dashed
line), and a solution of a 1:1 mixture of Fe(II) and 3,5-diCl-Sal-AHA
(thin blue solid line).
quantum yield of Fe reduction was calculated using the equation ΦFe
=
[Fe2+] × NA/(I × Δt). Ligand cleavage was monitored using circular
dichroism spectroscopy. The α-carbon of the α-hydroxy acid group is
chiral, providing a CD signal for the intact complex. Loss of the chiral
center upon cleavage is monitored as a decrease in the CD intensity.
This experiment is done under aerobic conditions, to allow air
oxidation of the reduced Fe so that the intact ligand will re-form the
cluster and provide CD intensity that accurately corresponds to the
amount of remaining ligand. The quantum yield for ligand cleavage
was calculated from the following equation: ΦL = [L] × NA/(I × Δt)
where [L] is the decrease in concentration of chiral ligand during
irradiation.
similar to the CD of a solution of the 3,5-diCl-Sal-AHA chelate
prepared with one equivalent of Fe(II)(NO3)2. In order to
determine the percentage of intact chelate represented by this
CD signal, the sample (without BPDS) was opened to the air in
the dark to allow the iron to air oxidize to Fe(III) under
nonphotolyzing conditions. This should result in reassembly of
the [Fe3(3,5-diCl-Sal-AHA)3(OCH3)]− cluster from any
remaining intact Sal-AHA chelate so that the CD signal
intensity can be directly compared to that of the initial solution
prior to irradiation. The CD spectrum in the presence of
Fe(III) corresponds to 53% of the chelate remaining intact after
irradiation, based on three independent measurements, or 47%
cleavage of the chelate. These data are consistent with a 2:1
ratio of iron reduction to chelate cleavage in the anaerobic
photolysis reaction.
While the 2:1 ratio determined above is consistent with that
observed for the Fe(III)-citrate complex,22 which has been
shown to be a Fe2(cit)2 dimer,23 the ratio is assumed to be 1:1
for the marine siderophores that bind a single iron. In order to
determine whether the photochemical reaction would still
occur with only a single Fe(III) in the complex, irradiation of a
solution that includes mixed-metal Fe/Ga complexes of 3,5-
diCl-Sal-AHA with the same trimeric structure as [Fe3(3,5-diCl-
Sal-AHA)3(OCH3)]−, was examined. The Ga(III) complex of
3,5-diCl-Sal-AHA3− has been structurally characterized.24 It has
a very similar structure to [Fe3(3,5-diCl-Sal-AHA)3(OCH3)]−,
but it is not photoactive, as expected. When Ga(III) and Fe(III)
are together combined with 3,5-diCl-Sal-AHA3−, a mixture that
includes the all-Fe and all-Ga trimers, as well as the 2:1 Fe:Ga
and 1:2 Fe:Ga mixed-metal trimers, is produced.24 This
facilitates an experiment to determine whether two Fe(III)
ions are required in the cluster for the photolysis reaction to
occur. The main features in the ESI mass spectrum of the
solution resulting from combination of Fe(III), Ga(III), and
3,5-diCl-Sal-AHA3− are centered near m/z of 1065, 1079, 1093,
and 1107. These features correspond to the Fe3, Fe2Ga, FeGa2,
and Ga3 clusters, respectively. Upon irradiation for 30 min
under either aerobic or anaerobic conditions, all three of the
Fe-containing peaks disappear, leaving only the Ga3 cluster, as
RESULTS
■
Metal:Chelate Ratios in Photolysis Reaction. Upon
irradiation of [Fe3(3,5-diCl-Sal-AHA)3(OCH3)]− with ultra-
violet light, the Fe(III) is reduced to Fe(II), and the 3,5-diCl-
Sal-AHA3− chelate is cleaved. (The nature of this cleavage is
addressed below.) This photolysis reaction can be monitored
from the point of view of either the iron reduction or the
chelate cleavage. Production of Fe(II) can be monitored
spectrophotometrically by addition of the phenanthroline
derivative BPDS, which shows an intense visible absorption
when coordinated to Fe(II) as Fe(BPDS)34−. The α-carbon of
the AHA group in the 3,5-diCl-Sal-AHA3− chelate is chiral, so it
shows a strong circular dichroism (CD) signal whether it is free
or coordinated to a metal. The photolysis reaction results in
loss of the chiral center, so the chelate cleavage can be
monitored by loss of the CD signal. Under aerobic conditions,
the CD signal disappears entirely after sufficient irradiation to
drive the photolysis reaction to completion, and a majority of
the iron (around 70%) is detected as Fe(II) based on the
absorbance at 535 nm when BPDS is present.10 Under
anaerobic conditions, however, while nearly all of the iron is
detected as Fe(II), some CD signal remains, indicating that
some of the chelate remains intact even after extensive
irradiation.
The ratio of Fe reduction to chelate cleavage was determined
for anaerobic irradiation of [Fe3(3,5-diCl-Sal-AHA)3(OCH3)]−,
so that reoxidation of Fe(II) by air is not available to drive
chelate photolysis beyond the single photolysis event ratio. As
indicated above, after 60 min of irradiation under UVA light in
5783
dx.doi.org/10.1021/ic500635q | Inorg. Chem. 2014, 53, 5781−5787

Inorganic Chemistry
Article
shown in Figure 2. After shorter periods of irradiation, the rate
of disappearance of all three Fe-containing clusters appears to
be similar (Figure S6).
irradiation, Φ, should correspond to reduction of two Fe(III)
with cleavage of one chelate, or Φ = ΦL = 1/2ΦFe. The quantum
yields determined by each method are reported in Table 1 for
Fe(III) complexes of four different phenyl ring-substituted
derivatives of Sal-AHA.
Photolysis Products of [Fe3(3,5-diCl-Sal-AHA)3-
(OCH3)]−. Irradiation of a methanolic solution of Na[Fe3(3,5-
diCl-Sal-AHA)3(OCH3)] by near-UV light under anaerobic
conditions for 60 min results in loss of the feature
corresponding to the monoanionic, trimeric cluster at m/z =
1065 in the ESI-MS experiment (Figure S1), with a feature at
m/z = 244 growing in (Figure S7). This new feature
corresponds to a loss of 44 mass units from the intact [3,5-
diCl-Sal-AHA]3− chelate, along with loss of the strong Fe
binding. This is consistent with decarboxylation of the AHA
moiety to produce an aldehyde with one fewer carbon atom.
Additional irradiation of this solution under anaerobic
conditions does not result in further changes. In order to test
whether this product acts like an aldehyde, propylamine was
added to the solution of the photolysis product. This resulted in
new features in the ESI-MS at m/z = 114 and 232 (Figure S8).
The m/z = 114 feature corresponds to the Schiff base that
would be formed by reaction of propylamine with 3-
aminopropionaldehyde, while the m/z = 232 feature corre-
sponds to the Schiff base of propylamine with 3,5-diCl-
salicylaldehyde.
Figure 2. ESI mass spectrum of Fe(III):Ga:(3,5-diCl-Sal-AHA) 1:2:3
in methanol (A) before irradiation and (B) after irradiation with UVA
light. The 1:2 ratio of Fe:Ga was used to compensate for the greater
sensitivity of the mass spectrometer to the Fe complexes; however the
same result is observed for the 1:1:2 ratio of Fe:Ga:L (see Figure S5).
Irradiation by near-UV light of a methanolic solution of
Na[Fe3(3,5-diCl-Sal-AHA)3(OCH3)] prepared in the same way
but under air results not in the aldehyde but rather in a different
product. The main product observed by ESI-MS under these
conditions occurs at m/z = 216 (Figure S9) and corresponds to
loss of 72 mass units from 3,5-diCl-Sal-AHA or a further 28
mass units from the aldehyde product of anaerobic irradiation.
A series of experiments was undertaken to determine the
conditions under which irradiation resulted in the aldehyde
versus the more extensive chelate cleavage. Conditions included
presence or absence of air, as well as exposure to air after
anaerobic irradiation. The results of these experiments,
monitored by ESI-MS, are listed in Table 2. Under conditions
that prevent air oxidation of Fe(II), the major product is the
aldehyde at m/z = 244. As noted above, the major product due
to photolysis under air is at m/z = 216, corresponding to a net
loss of CO from the aldehyde. When Na[Fe3(3,5-diCl-Sal-
AHA)3(OCH3)] is irradiated under anaerobic conditions, and
the product is then exposed to air in the dark, different
products appear at m/z = 576 and 732 (Figure S10). These
features correspond to Fe complexes that include the
carboxylate (designated L′) resulting from air oxidation of
the aldehyde product of anaerobic irradiation. During
irradiation under air in the presence of phenanthroline, which
partially protects Fe(II) from air oxidation back to Fe(III),
oxidation of the aldehyde is slowed substantially (Figure S11),
with a significant amount of aldehyde remaining after 2 h of
aerobic irradiation. Taken together, these data suggest the
reaction sequence shown in Scheme 2.
Quantum Yields for [Fe3(X-Sal-AHA)3(OCH3)]−. The
quantum yields for photolysis of the [Fe3(X-Sal-AHA)3-
(OCH3)]− complexes can be determined based on either
iron reduction or chelate cleavage. The quantum yield for
anaerobic iron reduction (ΦFe) is based on determination of
Fe(II) concentration through the spectrophotometric assay
with BPDS described above and shown in Figure 3. The value
Figure 3. UV/vis absorbance spectra of Na[Fe3(3,5-diCl-Sal-AHA)3-
(OCH3)] (0.0375 mM in methanol) in the presence of excess BPDS
after irradiation with UVA light for 0, 1, 3, 5, 7, 10, 20, and 30 min
(absorbance increasing with time). Inset: Absorbance at 535 nm, with
the points used to calculate quantum yields indicated by red arrows.
In order to test this proposed reaction scheme, the
compound with an m/z of 260, L′ = N-(3,5-dichlorosalicyli-
dene-β-alanine), was synthesized. While crystal structures of
this ligand with several different metals have been reported,25,26
the structure of the Fe complex has not. When L′ is added to
Fe(III), this produces the complex Fe(III)L′2, with a feature at
m/z = 576 in the ESI-MS experiment (Figure S4). This
corresponds to the feature observed in the mass spectrum of
of ΦFe is calculated to be somewhat lower when determined
under air presumably due to some competition between air
oxidation of the Fe(II) to Fe(III) and its formation of
Fe(II)(BPDS)34−. The quantum yield for ligand cleavage (ΦL)
is based on decrease in the concentration of the intact AHA
ligand determined by loss of the CD signal. The Fe reduction
to ligand cleavage stoichiometry determined above indicates
that the quantum yield of a single photochemical event upon
5784
dx.doi.org/10.1021/ic500635q | Inorg. Chem. 2014, 53, 5781−5787

Inorganic Chemistry
Article
a
Table 1. Quantum Yields for Irradiation of the [Fe3(X-Sal-AHA)3(μ3-OCH3)]− Complexes with Different Wavelength Ranges
UVB
ΦL
UVA
ΦL
LZC-420
X
ΦFe
Φ
ΦFe
Φ
ΦFe
ΦL
Φ
5-NO2
3,5-diCl
all-H
4.3 0.1
2.4 0.3
1.0 0.3
1.6 0.1
1.3 0.7
1.3 0.2
0.7 0.1
0.7 0.1
1.7
1.2
0.6
0.7
4.2 0.1
2.6 0.3
0.9 0.2
1.4 0.2
1.5 0.1
1.0 0.2
0.6 0.1
0.8 0.2
1.8
1.1
0.5
0.8
2.2 0.1
0.8 0.1
0.6 0.2
0.7 0.1
0.8 0.2
0.9
0.4
0.3
0.3
0.3 0.1
0.2 0.1
0.2 0.1
3-OCH3
a
1
Φ is the average of the values for ΦL and /2ΦFe, given in % yield.
Comparison of the amount of Fe(II) produced and chelate
cleavage incurred upon anaerobic irradiation, as monitored by
the spectrophotometric detection of the resulting Fe(II)-
(BPDS)3 complex, and the loss of the chiral center in the
AHA moiety monitored by the decrease in the CD signal,
respectively, demonstrates a 2:1 metal reduction to chelate
decarboxylation ratio. This is consistent with the ratio
suggested previously,10 based on the residual CD intensity
observed upon extensive irradiation only under anaerobic
conditions, and also the reported photochemical reaction of
Fe(III)2(citrate)2.22 This ratio is consistent with the pair of one-
electron reductions of Fe(III) to Fe(II) corresponding to a
single two-electron oxidative decarboxylation of the AHA
moiety to give the aldehyde.
Table 2. Species Observed by ESI-MS under Different
Conditions Due to Irradiation by UVA
dark air or
Ar
UVA
Ar
UVA
air
UVA Air +
phen
UVA Ar, then dark
air
m/z
1065
244
216
576
×
×
×
×
×
×
Scheme 2. Reaction Sequence for Na[Fe3(3,5-diCl-Sal-
AHA)3(OCH3)] Following Irradiation
If the 2:1 ratio of Fe reduction to chelate cleavage were the
result of a requirement for two reducible irons in the complex in
order to result in photochemically induced oxidative decarbox-
ylation of the AHA, then a cluster with only one Fe(III) would
not be expected to undergo photolysis. This is contrary to the
observation for the mixed FeGa clusters, in which not only the
Fe2Ga cluster but also the FeGa2 cluster is photoactive. While
the all-Ga cluster is photochemically inert, this experiment
shows that only a single Fe(III) is required for the
photochemical reaction to occur. These apparently contra-
dictory observations are rectified by considering results of laser
flash photolysis studies by Glebov at al. of other related Fe(III)-
carboxylate complexes. Those studies, under conditions in
which only a single Fe(III) is coordinated to the activated
carboxylates, were interpreted as indicating that “...[the
photolysis] mechanism is based on the formation of a long-
lived complex between Fe(II) and organic radical”.27 In the
presence of a second Fe(III), this Fe(II)-radical complex would
be further oxidized by one electron, resulting in the observed
two Fe(II) ions and the two-electron-oxidized decarboxylation
product. In the case of the α-hydroxy acids, that product would
be the aldehyde. In the absence of the second Fe(III), as in the
FeGa2 cluster, or the marine siderophores, the second electron
would go to any available site that would lower the overall
energy, including perhaps solvent.
Quantum yields for the photolysis of several members of the
[Fe(III)3(X-Sal-AHA)3(μ-OCH3)]− series were determined by
monitoring both the Fe(II) production and the loss of the
chiral center on the AHA group. Quantum yields based on the
amount of Fe(II) produced (ΦFe) are about twice those based
on the amount of loss of the AHA chiral center (ΦL). This 2:1
ratio of ΦFe:ΦL determined after brief exposure to light further
supports the ratio determined by measuring the amount of
Fe(II) produced and intact AHA remaining after exhaustive
photolysis as described above. The trend in the quantum yields
in each of the wavelength ranges examined agrees with the
relative rates of photolysis observed previously under sun-
light.10 That is, more electron-withdrawing ring substituents on
the phenolate of the X-Sal-AHA chelate result in higher
the product of reaction 2 in Scheme 2. Irradiation of this
complex results in a product with m/z = 216 (Figure S12),
identical to the final product observed after irradiation of
Na[Fe3(3,5-diCl-Sal-AHA)3(OCH3)] in the air. In contrast,
irradiation of L′ under Ar in the presence of Fe(II) does not
produce the feature at m/z = 216 (Figure S13).
DISCUSSION
■
The X-Sal-AHA series of chelates that incorporate a salicylidene
moiety and an α-hydroxy acid moiety, with variable electron-
donating or -withdrawing groups on the phenolate ring, was
reported to form trimeric clusters with Fe(III) in the form
[Fe3(X-Sal-AHA)3(μ3-OCH3)]−.10 These clusters were shown
to be photoactive in sunlight, with the resulting photochemical
reaction substantially decreasing the high stability constants of
the initial trimers, allowing preferential binding of the resulting
Fe(II) by a phenanthroline derivative. This irreversible
photochemical reaction was shown to result from the
interaction between the Fe(III) and the chelate, as there was
no observable change in the free chelate when irradiated by
sunlight. In this paper, details of that photochemical reaction of
the Fe(III) cluster complex are reported. The results of these
studies have important implications for the photochemical
reactions of the marine siderophores that inspired the X-Sal-
AHA chelates, as discussed below.
5785
dx.doi.org/10.1021/ic500635q | Inorg. Chem. 2014, 53, 5781−5787

Inorganic Chemistry
Article
quantum yields. In the UVA and UVB wavelength ranges, the
observed; however, the reported studies on photolysis of
marine siderophores all appear to have been conducted in air.
The data reported here provide significant insight into the
chemical basis for the difference in photolysis products for the
citrate-based and β-hydroxyaspartate-based marine sidero-
phores. The question as to why some marine bacteria evolved
to produce siderophores that give an oxygen-sensitive
photolysis product that results in cleavage of the fatty acid
moiety, while others evolved to produce siderophores that are
resistant to this cleavage, remains an interesting question.
quantum yields for the different complexes are in the range Φ =
1
ΦL = /2ΦFe = 0.5−2%.
Although exhaustive photolysis of Na[Fe3(3,5-diCl-Sal-
AHA)3(OCH3)] appears to produce different products depend-
ing on the presence or absence of air, the reaction under either
condition proceeds through decarboxylation of the AHA
moiety to produce an aldehyde. This step of the sequence of
reactions in Scheme 2 is likely initiated by a ligand-to-Fe(III)
charge transfer that is consistent with the excited state observed
for other Fe-carboxylate complexes in earlier laser flash
photolysis experiments.27 An electron transfer from the radical
produced by the initial photochemical event to a second Fe(III)
in the cluster results in oxidative release of CO2 to form the
aldehyde. In the absence of air, no further reaction takes place
and the aldehyde is the final observed product. The addition of
propylamine confirms the chemical nature of the aldehyde.
Although the Schiff base corresponding to reaction of
propylamine with the aldehyde product shown in reaction 1
of Scheme 2 is not observed, two products consistent with
displacement of 3-aminopropionaldehyde from the salicylidene
linkage by propylamine are observed. One is the Schiff base
resulting from reaction of propylamine with the salicylidene
fragment, and the other is the Schiff base resulting from
reaction of propylamine with the 3-aminopropionaldehyde
fragment. This latter species demonstrates the aldehyde nature
of the photolysis product.
ASSOCIATED CONTENT
* Supporting Information
■
S
Details of the synthesis and characterization of N-salicylidene-
β-alanine and its Fe(III) complex, details of the preparation of
the mixed Fe/Ga clusters, and Figures S1−S13 including all
mass spectra referred to in the text. This material is available
free of charge via the Internet at http://pubs.acs.org.
AUTHOR INFORMATION
Corresponding Author
*E-mail: Michael.baldwin@uc.edu.
■
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
This research was supported by NSF grant CHE-0955603. The
authors thank Prof. Apryll Stalcup for the use of the circular
dichroism spectrophotometer.
When the anaerobic product of photolysis, including the
aldehyde and Fe(II), are exposed to air without further
irradiation, the carboxylate product of reaction 2 in Scheme 2
(designated L′) is produced. When phenanthroline is present
to protect the Fe(II) from air oxidation to Fe(III), much less
oxidation of the aldehyde product occurs. Together, these
observations show that the aldehyde oxidation step is
dependent on both O2 and iron, but is not light driven. The
final step to give the product of reaction 3 in Scheme 2 is light
driven, as the reaction stops at the carboxylate in the absence of
further irradiation. It is also dependent on Fe(III), as the
synthetic L′ does not undergo photolysis in the absence of iron
or in the presence of only Fe(II).
The comparison of aerobic and anaerobic photolysis
products of Na[Fe3(3,5-diCl-Sal-AHA)3(OCH3)] has impor-
tant implications for the differences observed for marine
siderophores with different AHA frameworks. Those based on a
citrate backbone are reported to undergo simple decarbox-
ylation, while the β-hydroxyaspartate-based siderophores
undergo much more extensive cleavage.8,9 The studies reported
here suggest that this apparently large difference in reactivity is
not due to differences in the initial photochemical and chemical
steps. Rather, the difference is likely due to differences in
susceptibility of the decarboxylation product to air oxidation in
the presence of iron. Decarboxylation of the citrate moiety
produces a ketone, which is resistant to air oxidation. However,
decarboxylation of the β-hydroxyaspartate moiety produces an
aldehyde, which, in the presence of iron and oxygen, is oxidized
to a carboxylate, which can undergo further photolysis when
coordinated to Fe(III). In many of the amphiphilic marine
siderophores, the characteristic fatty acid group is attached near
the photolysis site and is cleaved from the main body of the
molecule when the additional photolysis reactions occur. If the
β-hydroxyaspartate-based marine siderophores were irradiated
under anaerobic conditions, this would predict that simple
decarboxylation of the AHA group to the aldehyde would be
REFERENCES
■
(1) Martin, J. H.; Coale, K. H.; Johnson, K. S.; Fitzwater, S. E.;
Gordon, R. M.; Tanner, S. J.; Hunter, C. N.; Elrod, V. A.; Nowicki, J.
L.; Coley, T. L.; Barber, R. T.; Lindley, S.; Watson, A. J.; VanScoy, K.;
Law, C. S.; Liddicoat, M. I.; Ling, R.; Stanton, T.; Stockel, J.; COllins,
C.; Anderson, A.; Bidigare, R.; Ondrusek, M.; Latasa, M.; Millero, F. J.;
Lee, K.; Yao, W.; Zhang, J. Z.; Friederich, G.; Sakamoto, C.; Chavez,
F.; Buck, K.; Kolber, Z.; Greene, R.; Falkowski, P.; Chisholm, S. W.;
Hoge, F.; Swift, R.; Yngel, J.; Turner, S.; Nightingale, P.; Hatton, A.;
Liss, P.; Tindale, N. W. Nature 1994, 371, 123−129.
(2) Bruland, K. W.; Donat, J. R.; Hutchins, D. A. Limnol. Oceanogr.
1991, 36, 1555−1577.
(3) Butler, A.; Theisen, R. M. Coord. Chem. Rev. 2010, 254, 288−296.
(4) Holt, P. D.; Reid, R. R.; Lewis, B. L.; Luther, G. W., III; Butler, A.
Inorg. Chem. 2005, 44, 7671−7677.
(5) Zhang, G.; Amin, S. A.; Kupper, F. C.; Holt, P. D.; Carrano, C. J.;
Butler, A. Inorg. Chem. 2009, 48, 11466−11473.
(6) Amin, S. A.; Green, D. H.; Kupper, F. C.; Carrano, C. J. Inorg.
Chem. 2009, 48, 11451−11458.
(7) Barbeau, K.; Rue, E. L.; Trick, C. G.; Bruland, K. W.; Butler, A.
Limnol. Oceanogr. 2003, 48, 1069−1078.
(8) Barbeau, K.; Zhang, G.; Live, D. H.; Butler, A. J. Am. Chem. Soc.
2002, 124, 378−379.
(9) Barbeau, K.; Rue, E. L.; Bruland, K. W.; Butler, A. Nature 2001,
413, 409−413.
(10) Sayre, H.; Milos, K.; Goldcamp, M. J.; Schroll, C. A.; Krause, J.
A.; Baldwin, M. J. Inorg. Chem. 2010, 49, 4433−4439.
(11) Rosell, F. I.; Mauk, A. G. Coord. Chem. Rev. 2011, 255, 737−
756.
(12) Kennedy, D. P.; Incarvito, C. D.; Burdette, S. C. Inorg. Chem.
2010, 49, 916−923.
(13) Ciesienski, K. L.; Haas, K. L.; Dickens, M. G.; Tesema, Y. T.;
Franz, K. J. J. Am. Chem. Soc. 2008, 130, 12246−12247.
(14) Ellis-Davies, G. C. R.; Kaplan, J. H. Proc. Natl. Acad. Sci. U.S.A.
1994, 91, 187−191.
5786
dx.doi.org/10.1021/ic500635q | Inorg. Chem. 2014, 53, 5781−5787

Inorganic Chemistry
Article
(15) Bandara, H. M. D.; Kennedy, D. P.; Akin, E.; Incarvito, C. D.;
Burdette, S. C. Inorg. Chem. 2009, 48, 8445−8455.
(16) Kaplan, J. H.; Ellis-Davies, G. C. R. Proc. Natl. Acad. Sci. U.S.A.
1988, 85, 6571−6575.
(17) Kupper, F. C.; Carrano, C. J.; Kuhn, J.-U.; Butler, A. Inorg.
Chem. 2006, 45, 6028−6033.
(18) Martin, J. D.; Ito, Y.; Homann, V. V.; Haygood, M. G.; Butler, A.
J. Biol. Inorg. Chem. 2006, 11, 633−641.
(19) Ito, Y.; Butler, A. Limnol. Oceanogr. 2005, 50, 1918−1923.
(20) Parker, C. A. Proc. R. Soc. London, Ser. A 1953, 220.
(21) Nyhus, K. J.; Wilborn, A. T.; Jacobson, E. S. Infect. Immun. 1997,
434−438.
(22) Abrahamson, H. B.; Rezvani, A. B.; Brushmiller, J. G. Inorg.
Chim. Acta 1994, 226, 117−127.
(23) Shweky, I.; Bino, A.; Goldberg, D. P.; Lippard, S. J. Inorg. Chem.
1994, 33, 5161−5162.
(24) Chrisman, M. A.; Grabo, J. E.; Krause, J. A.; Baldwin, M. J.,
manuscript in preparation.
(25) Chandra, S. K.; Basu, P.; Ray, D.; Pal, S.; Chakravorty, A. Inorg.
Chem. 1990, 29, 2423−2428.
(26) Plesch, G.; Kettmann, V.; Sivy, J.; Svajlenova, O.; Friebel, C.
Polyhedron 1998, 17, 539−545.
(27) Glebov, E. M.; Pozdnyakov, I. P.; Grivin, V. P.; Plyusnin, V. F.;
Zhang, X.; F. Wu, N. D. Photochem. Photobiol. Sci. 2011, 10, 425−430.
5787
dx.doi.org/10.1021/ic500635q | Inorg. Chem. 2014, 53, 5781−5787