The synthesis and characterization of a side-by-side iron phthalocyanine dimer

Journal of Porphyrins and Phthalocyanines

J. Porphyrins Phthalocyanines 2011; 15: 277–292

DOI: 10.1142/S1088424611003264

Published at http://www.worldscinet.com/jpp/

The synthesis and characterization of a side-by-side iron

phthalocyanine dimer

Wei He and Marya Lieberman*^

Department of Chemistry and Biochemistry, University of Notre Dame, Notre Dame, IN 46556, USA

Received 30 March 2011

Accepted 1 May 2011

ABStrAct: The QCA paradigm is one of the approaches to decrease the size scale of computing

devices. When molecules are used as QCA cells, they may be able to perform computing at room

temperature. This paper describes a novel molecular QCA cell candidate which is a side-by-side iron

phthalocyanine dimer, and an investigation of its optical and redox properties. The new dodeca(pentyloxy)

substituted side-by-side iron phthalocyanine dimer, along with the octa(pentyloxy) iron phthalocyanine

monomer, are soluble in non-polar organic solvents. These compounds were isolated by gel permeation

chromatography (GPC) and high-performance liquid chromatography (HPLC) to ﬁnal purities of 98%

and 99%, respectively. The NMR spectra of both compounds in CDCl₃are broad due to aggregation,

but become well resolved after the addition of the coordinating solvent pyridine-d₅. Addition of pyridine

also gives changes in the UV-vis spectra and electrochemical peaks of both monomer and dimer in

dichloromethane indicative of axial iron coordination. The electrochemical data indicates the loss of

pyridine ligands from the oxidized products of both monomer and dimer. The comproportionation

constant of side-by-side phthalocyanine dimer shows that its oxidized and reduced mixed-valence

complexes are fairly stable. The dimer is thus a candidate for molecular QCA systems.

KEYWordS: iron phthalocyanine, Fe phthalocyanine, pc, dimer, aggregation, HPLC, electrochemistry,

UV-vis, NMR, mass spectrometry, pyridine.

In chemical terms, the smallest possible quantum dot

IntroductIon

can be thought of as a redox center in a suitable mol-

The quantum-dot cellular automata (QCA) concept is

ecule. There are many examples of molecules that con-

a paradigm for transmitting and processing information

tain arrays of redox centers in precise locations, and

in arrays of quantum dots. The positions of charges in the

calculations have shown that a dimeric mixed-valence

quantum dot array are used to perform logical computa-

compound (containing two redox centers connected by

tions. According to the QCA theory [1, 2], the smaller

a bridging moiety to mediate electron transfer) could be

the quantum dots are, the better they can work as charge

used as an individual QCA cell [1]. Our ultimate goal

containers; when the cell dimensions approach molecu-

is to self-assemble molecular QCA cells onto substrates

lar sizes, it is predicted that room temperature operation,

to perform computing. Since self-assembly methods

very high speed, and low energy consumption can all be

have shown the capability for patterning nanostructures

achieved [3]. In the QCA paradigm, each cell is ﬁeld-

[4–7], and electron beam lithography is capable of mak-

coupled to its neighbors so that individual contacts to the

ing sub-10 nm size structures [8], it may be possible to

cells are not necessary, which is particularly attractive for

employ lithography and self-assembly methods to help

molecular electronics.

pattern QCA molecules on the substrates. This paper

describes the synthesis and characterization of a pc dimer

which is well suited to both QCA function and surface

assembly.

^SPP full member in good standing

Phthalocyanine compounds have, since their ﬁrst dis-

covery in 1907 and the elucidation of their structure in

*Correspondence to: Marya Lieberman, email: mlieberm@

nd.edu, tel: +1 574-631-4665, fax: +1 574-631-6652

278

W. He anD M. LIeberMan

1934, achieved a notable success as a model structure for

experimental organic, theoretical, and physical chemi-

cal study. For many years the insolubility of unsubsti-

tuted phthalocyanines hindered studies on their reactivity

and properties. The poor solubility results from strong

π-stacking interactions between the phthalocyanine rings

and weak solvent-phthalocyanine interactions. Kenney’s

group found that the axial Si-OH bonds in (OH)₂SiPc

could react with trialkylchlorosilanes to form bis-siloxy-

SiPcs with good solubility in organic solvents [9]. An

alternative way to enhance the solubility of pcs in organic

solvents was developed by Hanack, who introduced

peripheral substitutents into the macrocycle [10]. This

provides a convenient way to increase the solubility of

the compounds while freeing up the axial positions for

other roles.

In order to make phthalocyanine dimers, one can either

stack the phthalocyanines one atop the other [11–13], or

construct two phthalocyanine rings side-by-side [14–16].

Previous work in our group has explored thestacked dimer

geometry [17, 18], but subsequent calculations [19] sug-

gested that a side-by-side dimer would be more suitable

for our purposes in fabrication of molecular electronic

devices. In particular, the side-by-side dimer would be

more accessible for Ultra HighVacuum Scanning Tunnel-

ing Microscope (UHV-STM) studies, since both sides of

the molecule would be equally exposed atop a solid sur-

face. An iron phthalocyanine was chosen in part because

of the rich axial chemistry of the iron [20–22], which

could facilitate surface attachment on self-assembled

monolayers. A tetra-nitrilo benzene was chosen to form

the linker between the two phthalocyanine rings, because

the conjugated system facilitates electron delocalization.

As seen in molecule 2 in Fig. 1, the two cojoined phtha-

locyanine rings are expected to share this benzene ring,

with an expected distance of 1.2 nm between the two iron

atoms. The phthalocyanine rings were decorated with

pentyloxy groups in order to increase their solubilities;

this was an important synthetic goal both to facilitate

characterization of the products, and because surface

attachment requires high purity material.

There are generally two strategies for making unsym-

metrical phthalocyanines; the ring expansion method

and the statistical method. The ring-expansion method

utilizes boron subphthalocyanines (B-SubPcs) reacting

with diiminoisoindoline analogues [23, 24]. The ﬁnal

yields of unsymmetrical pcs varied from 8–20%. One of

the shortcomings of this method is the thermal instabil-

ity of B-SubPc [25], and in our hands, ring expansion of

B-SubPc was not sucessful for generation of the desired

FePc dimers.

The other method to make unsymmetrical phthalo-

cyanines is based on condensation of two different ortho-

dinitriles, giving statistical mixtures of products which

must then be puriﬁed [26]. Reaction conditions are simi-

lar to standard phthalocyanine syntheses: high reaction

temperature (> 150 °C), long heating time (> 24 h), and

an inert environment (argon or nitrogen gas). 1-chlo-

ronaphthalene or N,N-dimethylaminoethanol (DMAE)

is typically used as the solvent. The reported yields of

unsymmetrical pcs vary from 1% (unsymmetrical CoPc)

[27] to 10% (unsymmetrical ZnPc) [28] to 19% (unsym-

metrical NiPc) [29].

Puriﬁcation of phthalocyanine products is often dif-

ﬁcult, yet purity is critical for molecules that will even-

tually be used for surface deposition studies. Extensive

solvent washing and sublimation are commonly used to

remove impurities from the relatively insoluble unsub-

stituted phthalocyanines [30]. For phthalocyanine deriva-

tives with higher molecular weight, the sublimation

method is not applicable because peripheral substitutents

of the phthalocyanine often fragment during heating. TLC

and chromatography on silica are often used to separate

phthalocyanine fractions, but the separation that can be

achieved is limited by the relatively low number of theo-

retical plates. Gel permeation chromatography separates

R

N

Fe

N

Fe

N

Fe

N

R

N

R

1

2

Fig. 1. Molecular structure of iron phthalocyanine monomer 1 and side-by-side iron phthalocyanine dimer 2

J. Porphyrins Phthalocyanines 2011; 15: 278–292

THe SynTHeSIS anD CHaraCTerIzaTIOn Of a SIDe-by-SIDe IrOn PHTHaLOCyanIne DIMer

279

materials on the basis of molecular size, which is useful

Bis-diiminoisoindoline (8). To make the linker for

the phthalocyanine dimer, the commercially available

1,2,4,5-tetracyanobenzene (7) was bubbled with gaseous

ammonia in the presence of sodium methoxide and meth-

anol for 45 min to form bisdiiminoisoindoline (8). 8 was

insoluble in methanol and precipitated out, giving a yield

of 99%. IR was used to conﬁrm the disappearance of the

C≡N stretching frequency at 2228 cm^-1.

for removing starting materials from crude pc products

and separating products of different molecular weights

[31–34]. However, the separation obtained is limited.

We turned to high-performance liquid chromatography

(HPLC) [35] to achieve better phthalocyanine purity for

surface deposition. As a side beneﬁt, the diode-array

detector allowed us to identify each fraction as it came

out of the stationary phase of the column. 2 was obtained

in 12% yield based on 1,2,4,5-tetranitrilobenzene, and

analytical HPLC result showed its purity at 98%.

Statistical synthesis of 1 and 2

The ring-expansion [24] of boron sub-phthalocyanine

was ﬁrst attempted to synthesize phthalocyanine dim-

ers, but in our hands it yielded little isolable phthalocya-

nine product. We believe that the self-condensation of

the bisdiiminoisoindoline and degradation of the boron

(hexapentyloxy) sub-phthalocyanine at high tempera-

tures hindered the formation of the desired phthalocya-

nine dimers. The statistical method as shown in Scheme

2 was used to make the pc dimer compound 2 with a

yield of 12% based on tetracyananobenzene. One of the

major challenges in this approach was to overcome the

poor solubility of bisdiiminoisoindoline (8). To solve

this problem, bisdiiminoisoindoline was ﬁrst heated in

DMAE to boiling point. FeCl₂·4H₂O and a DMAE solu-

tion of 6 were then added drop-wise from different necks

of a 3-neck ﬂask over half an hour. The reaction was done

under a blanket of NH₃gas. This procedure was intended

to keep the concentration of 6 low and thus increase the

chance for condensation with 8 to form the pc dimer 2.

Still, the pc monomer 1 was the major product and it was

hard to separate it from the pc dimer.

After 72h, the reaction was stopped and the solid was

placed in a Soxhlet extractor and washed for > 24 h each

with acetone and methanol to remove impurities. The

crude green phthalocyanine fraction, composed of 1 and

2, was extracted with chloroform until the extract in the

extractor thimble was nearly colorless. The crude yield

was about 300 mg. After gel-permeation chromatography

(GPC) and HPLC, 98% pure 2 (based on the HPLC chro-

matography data) with a yield of 12% (from the limiting

agent of tetracyanobenzene) was obtained. Our attempts

at forming the side-by-side pc dimer were similar to the

procedure of Lelievre’s work [16], although the yield

from the limiting agent in his report was 2.4%. The

main differences in our procedure were pre-heating

the bisdiiminoisoindoline (8) in DMAE before adding

the other ingredients, and carrying out the entire reaction

under a blanket of ammonia gas. Kobayashi’s group [33]

reported a 12% yield for the synthesis of his dinucleating

pc ligand, non-metalated, by statistical synthesis in

DMA, and about 8% each for the Co₂and Zn₂metalated

analogs of 2.

With pure iron octakis(pentyloxy)phthalocyanine

monomer (1) and side-by-side bis iron dodeca(pentyloxy)

phthalocyanine dimer (2) (as seen in Fig. 1) in hand,

detailed spectroscopic characterization and electrochem-

ical measurements were obtained in organic solvents. We

were particularly interested in the effect a strongly coor-

dinating solvent (pyridine) would have on the aggrega-

tion state of the phthalocyanines in solution, as we felt

that surface deposition would be more likely to yield iso-

lated molecules for STM analysis if the phthalocyanines

started out more or less monomeric in solution.

rESultS And dIScuSSIon

Synthesis

1,2-dicyano-4,5-bis(pentyloxy)benzene (6). The syn-

thesis route for 1,2-dicyano-4,5-bis(pentoxy)benzene is

shown in Scheme 1, which is modiﬁed from the work

done by Hanack [36]. This route starts from the bromi-

nation of catechol (3) [37]. The yield of the resulting

4,5-dibromocatechol (4) is 45%. O-alkylation of 4 with

1-bromopentane in DMF produces dialkyloxy dibromo-

catechol (5, 85%) in good yield in the presence of K₂CO₃

(Hanack used NaOCH₃as the base in this reaction). The

aromatic bromide (5) was then heated with CuCN in

DMF solution, and phthalonitrile (6, 59%) was obtained.

The overall yield of 6 from catechol was 23%. Copper

octa(pentyloxy)phthalocyanine was obtained as a side

product in the cyanization reaction.

OH

Br

OH

Br₂

acetic acid

45%

3

4

Br-C₅H₁₁/K₂CO₃

DMF

85%

Br

R

NC

R

CuCN

DMF

59%

R

R=pentyloxy

Puriﬁcation of 1 and 2

5

6

Gel permeation chromatography (using BioBeads

SX1, MW exclusion 600-14,000 amu) using toluene as

Scheme 1. The synthesis pathway to 1,2-dicyano-4,5-

bis(pentyloxy)-benzene (6)

J. Porphyrins Phthalocyanines 2011; 15: 279–292

280

W. He anD M. LIeberMan

R

N

R

N

Fe

N

R

N

R

N

6

+

R

N

OH

1 (29%)

FeCl₂·4H₂O

+

reflux for 72 hours

with NH₃bubbling

R

N

R

HN

NH

N

R

N Fe N

N

R

Fe

N

8

N

R

R=OC₅H₁₁

2 (12%)

Scheme 2. Synthesis of 1 and 2

eluent was used to enrich the fraction of dimer

in the crude phthalocyanine material. The

beads were swollen in toluene and packed

into a chromatography column. After passage

of the crude phthalocyanine material through

the column, the ﬁrst fraction was darker green

and this (about 1/3 of the crude material)

was the fraction used for HPLC puriﬁcation.

It consisted of about 50% 1 and 50% 2. The

second fraction coming out the column was

bright green and consisted almost entirely of

pc monomer (1, 96% purity before HPLC).

HPLC separation was conducted using a

semi-prep column of normal phase SiO₂and

Fig. 2. HPLC chromatogram of 2 (silica column, 2% py in CHCl₃). Solid

line in (b) is the UV-vis spectrum at the retention time of 5.7 min which was

proved to be 1, and dashed line is UV-vis spectrum at the retention time of

6.7 min which was proved to be 2

an isocratic eluent of 2% pyridine in chloroform. Two

major bands were present in the crude material, both with

intense absorbance spectra characteristic of phthalocya-

nines, and the later band showing multiple peaks con-

sistent with a dimeric pc. This band was assigned as the

dimer, and the mass spectra of the puriﬁed fractions later

conﬁrmed the assumption. Compounds 2 (Fig. 2a) and 1

(Fig. S2, see the Supplementary material) were obtained

in > 98% purity, as estimated by analytical HPLC. The

process is reasonably scalable; a total of 130 mg of 1 and

120 mg of 2 were eventually isolated and puriﬁed.

are fragile to the FAB ionization process, and sometimes

the molecular peak is nowhere to be found. Based on the

work of Lelievre [16], FAB usually breaks an alkyloxy

chain of the main frame of the phthalocyanine during

the ionization process. MALDI is a soft ionization tech-

nique used especially for fragile molecules. MALDI can

measure the m/z of a molecule with molecular weight up

to 10,000 Da [39–42] and it proved a reliable analytical

method for 1 and 2.

Compound 1 gave a strong peak at a m/z value of

1257.43 (theo. [M + H]⁺1257.99), and the phthalocya-

nine dimer 2 gave intact singly charged ions, at a m/z

value of 2092.79 (theo. [M + H]⁺2092.21). The error of

MALDI measurements is 0.1%, therefore for molecule

2, a 2 dalton deviation would be acceptable. In many of

the mass spectra, m/z peaks corresponding to ammonia

adducts were observed. After HPLC puriﬁcation, m/z

Mass spectra

Phthalocyanines are often ionized by methods such as

FAB, electrospray, andMALDI.Accordingtothereportof

Freas [38], the macrocyclic backbone of phthalocyanine

is difﬁcult to fragment, but substituted phthalocyanines

J. Porphyrins Phthalocyanines 2011; 15: 280–292

THe SynTHeSIS anD CHaraCTerIzaTIOn Of a SIDe-by-SIDe IrOn PHTHaLOCyanIne DIMer

281

peaks corresponding to pyridine adducts were found for

both 1 and 2.

Even in HPLC puriﬁed 2, which was 98% pure as

assayed by HPLC peak integration, a strong MALDI

peak of 1 or its pyridine coordination product was still

present. The use of MALDI for quantitative analysis is

limited [43–45]. The absolute intensity of the MALDI

signal depends on the laser power, the matrix crystalliza-

tion conditions, the ease of ionization of each compound,

and the point-to-point repeatability. Because puriﬁed 2

contained only about 2% 1, we think that 1 is much easier

to ionize than 2.

Fig. 4. Schematic diagram of electronic transitions of phthalo-

cyanine [51]

An m/z peak which equaled twice the molecular

weight of the iron phthalocyanine monomer was always

found in MALDI or FAB (see Fig. S6 in Supplementary

material). Since several groups [11, 46, 47] have reported

stacked metal pc dimers bonded via metal-metal bonds,

we believe those peaks may correspond to the Fe-Fe

bonded stacked phthalocyanine dimers; it is also possible

the pc rings could dimerize during the MS ionization

process. We see no evidence of this dimer species in the

HPLC of 1.

solvent so that very high concentration of pyridine was

achieved. Detailed data can be seen in the Supplementary

material.

Discussion of UV-vis spectra of 1. As illustrated in

Fig. 4, in the metal-free pc monomer, the symmetry is

lowered to D_2hfrom the D_4hsymmetry of the metallo pc

monomer because of the asymmetrical binding of the two

H atoms. The excited state is split into two states termed

Q_xand Q_y. These then are the two principal absorptions

in the mononuclear species [48–51].

uV-vis titration of 1 and 2 with pyridine

Because of the coordination capability of pyridine to

iron, one or two pyridine molecules could ligate to 1. For

molecule 2, one, two, three or four pyridines could ligate.

Pc molecules have large aromatic systems and their inter-

molecular π-π interactions are strong, so π-stacking and

aggregation are common. These two factors play important

roles in the changes of the UV-vis spectra of pc molecules.

Regular pc molecules tend to have cofacial aggregation,

as shown in phase A in Scheme 3 [13, 35, 52, 53]. When

pc molecules aggregate, the Q-band peak tends to shrink

and broaden and its shoulder peaks increase in intensity

[15, 54]. In Dominguez’s work, the UV-vis spectra of a

poly(ethyleneoxide)-capped phthalocyanine in CHCl₃,

which physically can only aggregate to a dimer, were taken.

One of the clearest ﬁndings was that the Q-bands shrank

drastically when the concentration of the pc was increased.

At the same time, the intensity of the shoulder peaks

increased. Their results showed that for a cofacial dimer

species, the Q-band was much

Results. Pure 1 (99%) and 2 (98%) fractions were

analyzed by UV-vis measurements. Their UV-vis spectra

did not show the characteristic pc features. For example,

the Q-band of both compounds was unusually weak. The

MALDI spectra of these materials showed py-coordi-

nated pcs were present, as described in the previous sec-

tion. Because of the axial coordination capabilities of 1

and 2, we thought the abnormal UV-vis results were from

pyridine coordination. Pyridine titration experiments

were performed on 1 and 2 (Fig. 3). First, the pyridine

residue was removed from HPLC-puriﬁed 1 and 2 under

high vacuum. Then, pyridine was injected into chloro-

form solutions of 1 and 2. Since the maximum volume

change due to the addition of pyridine was only 6%, the

concentrations of the pcs were considered to be constant.

After the titration experiments, two separate experi-

ments were carried out where pyridine was used as the

smaller than usual, comparable in

intensity to the shoulder peaks.

In the case of our experiments,

the concentration of pc remained

constant and the concentration of

pyridine changed. With excess

pyridine, the aggregates of 1

should break up and 1-(py)₂should

be formed in the end. Five possi-

ble phases (A′, A, B′, B and C) are

presented in Scheme 3. Three of

them (A, B, C) correspond to the

UV-vis spectra at the three indi-

cated stages (A, B, C) of Fig. 3a.

Fig. 3. UV-vis: pyridine titration. The arrows indicate the direction the spectra change

when pyridine concentration is increased. (a): 1(A: 1 in CHCl₃, B: 30.1:1 py/1 in CHCl₃,

C: 1 in py); (b): 2(A: 2 in CHCl₃, B: 181:1 py/2 in CHCl₃, C: 2 in pyridine. [1] = 6.4 ×

10^-3M, [2] = 3.1 × 10^-3M)

J. Porphyrins Phthalocyanines 2011; 15: 281–292

282

W. He anD M. LIeberMan

cofacial aggregation [35]. This assignment was supported

by the fact that the intensity of the peak at 700 nm in Fig.

S8a (see the Supplementary material) decreased when

the concentration of pyridine increased. The existence of

the peak at 700 nm, indicated that not all the molecules

of 1 were present as 1-(py)₂, even when neat pyridine was

used as the solvent.

There are two kinds of charge transfer peaks normally

seen in the UV-vis spectra of FePcs: one caused by MLCT

(metal-ligand charge transfer) and for alkoxyl-substituted

pcs, one caused by oxygen-to-pc(π*) [20, 60] charge

transfer. In supplementary Fig. S8a, the peak at 425 nm

remains at the same wavelength for the whole titration pro-

cess, so it is unlikely to be caused by MLCT. Therefore,

the peak at 425 nm was assigned to the oxygen-to-pc(π*)

charge transfer [17, 61]. Since the peak at 402 nm became

bigger when the concentration of pyridine increased, it

was reasonable to assign this peak as the MLCT peak.

Discussion of UV-vis spectra of 2. As shown in previ-

ous reports on side-by-side metal-free [62–64] and metal-

lophthalocyanine dimers [16, 33, 63, 65, 66] the Q-bands

of side-by-side pc dimers are blue-shifted compared to

the corresponding monomers. An exciton coupling model

explains the blue shift (Fig. 4). In the side-by-side dimer,

there are transition moments of the two excited states

(Q_x, Q_y) that couple, in and out-of-phase, between the

two halves of the binuclear molecule. In D_2hsymmetry of

side-by-side phthalocyanine dimer, the Q_xand Q_ystates

(as in D_2hsymmetry of metal free phthlocyanine mono-

mer) split and the coupling result in a pair of nondegener-

ate in-phase higher energy combinations (Q_y+, Q_x+) and a

pair of lower energy out-of-phase combinations (Q_y-, Q_x-)

(Fig. 4) [51]. The lower energy combinations (Q_y-, Q_x-)

give the red shifted Q-band.

Scheme 3. Reactions of 1 with pyridine

At the very beginning of the titration process a very

broad and suppressed Q-band was seen, and the Soret

band was suppressed too. These features are consistent

with extensive cofacial aggregation of 1 (stage A of

Fig. 3a) [35]. Because of the aggregation, the Q-band and

the shoulder peaks (673, 586 nm) were very broad.

As pyridine was added to the solution, the coordina-

tion of pyridine to the central Fe atom of 1 broke the

aggregation and 1-py was formed (phase B′ in Fig. 3a).

Aggregation (K_d12) could occur to form cofacial dimers

(phase B in Scheme 3). The new Q-band and shoulder

peak (668 nm and 585 nm) are assigned to this cofa-

cial dimer. The UV-vis spectrum of phase B (stage B in

Fig. 3a) showed broad peaks at 681 nm and 589 nm, and

the intensity of the Soret band (330 nm) was very small

as expected for a partially aggregated species.

At high pyridine concentrations, a intense Q-band

(660 nm, log ε = 4.12) was ﬁnally observed. Two shoul-

der peaks were found at 631 nm and 569 nm, which

are assigned as vibronic excitation [55–57]. A strong

Soret band (343 nm, log ε = 4.22) was also found. J.

Janczak has measured the absorption coefﬁcient for

FePc(pyrazine)₂in 1-chloronaphthalene (662 nm, log

ε = 6.71) which is signiﬁcantly larger than our observed

extinction coefﬁcient; however, this involved a different

axial ligand [57]. MJ Stillman also measured the absorp-

tion coefﬁcient for FePc(py)₂in DMSO (667 nm, log

ε = 4.88) [58]. Comparing this absorption coefﬁcient

with that of molecule 1 in pyridine (660 nm, log ε =

4.12), the difference is not substantial considering a dif-

ferent solvent was used. After a large excess of pyridine

was added, the equilibrium in the solution shifted to form

a six-coordinated Fe complex with both axial positions

occupied by pyridine (1-(py)₂, phase C in Scheme 3).

This broke up the cofacial dimer and isolated 1-(py)₂was

formed. The shoulder at 700 nm was assigned as residual

Again, axial coordination and co-facial aggregation

play important roles in the UV-vis spectra of 2. Five

possible phases (A′, A, B′, B and C) are presented in

Scheme 4. Three of them (A, B, C) are assigned to the

UV-vis spectra at three stages (A, B, C) in Fig. 3b. In the

Scheme 4. Reaction schemes of 2 with pyridine

J. Porphyrins Phthalocyanines 2011; 15: 282–292

THe SynTHeSIS anD CHaraCTerIzaTIOn Of a SIDe-by-SIDe IrOn PHTHaLOCyanIne DIMer

283

presence of excess pyridine, the UV-vis spectra indicate

that the aggregation of 2 is broken and a highly coordi-

nated species (2-(py)₄and/or 2-(py)₃) is formed.

resolved NMR spectra [21, 67–69] even for soluble pc

monomers and dimers which have side chains [54, 70,

71]. Because phthalocyanine molecules are big (> 2 nm)

and rigid, they have limited solubility and tend to aggre-

gate in solution, which decreases their tumbling rate.

Their aromatic cores have relatively long proton relax-

ation times. This behavior is also observed for metallo

pc molecules [27, 72, 73]. For typical NMR measure-

ments on pcs, low concentrations [27, 54, 73] and special

solvents, such as CS₂[16] or toluene [73], were used to

depress aggregation.

The NMR spectra of 1 and 2 were obtained in CDCl₃

but both spectra appeared to contain multiple species as

shown by the number and broadened envelopes of peaks.

In view of the UV-vis results, both phthalocyanines are

expected to be partially aggregated under these condi-

tions, and the aggregates are high enough in molecular

weight to cause NMR peak broadening due to slow tum-

bling rates. A trace amount of pyridine was present due

to its use in the HPLC eluent, so one or two pyridine

molecules could ligate to 1 and make mixtures of 1-py

and 1-(py)₂. For the dimeric phthalocyanine 2, the mix-

ture could include 2-py, 2-(py)₂, 2-(py)₃and even 2-(py)₄.

As seen in Fig. S10, because of the existence of multiple

coordination compounds, some strongly aggregated, and

multiple isomers of these compounds, their NMR spectra

were broad and very complicated. Addition of about four

drops of py-d₅clariﬁed the NMR spectra greatly for both

1 and 2. Figures 5 and 6 show NMR spectra of 1 and

2 in CDCl₃/py. With addition of excess pyridine, all the

coordination positions were occupied and single coordi-

nation compounds were formed. In the NMR tube reac-

tion, py-d₅was used. At this point, the NMR spectra of 1

and 2 were greatly simpliﬁed.

When no pyridine is present, the UV-vis spectrum

of 2 (stage A in Fig. 3b) shows a broadened Q-band

(683 nm) and a shoulder (582 nm). The low-energy peak

(745 nm) due to cofacial aggregation was fairly strong,

and the Soret band (370 nm) was weak. This spectrum

corresponds to the aggregated phase A in Scheme 4.

As pyridine is added, these aggregates are broken up as

mono, bi, or tripyridyl-2 molecules are formed by coordi-

nation to the two Fe atoms (phase B in Scheme 4). Some

aggregation appears to persist until a very high concentra-

tion of pyridine is reached. Phase B of Scheme 4 shows

cis-2 (Fe(py)-Fe(py)) aggregates. However, even for

a two-pyridine addition product, there also could be trans-

2(Fe(py)-Fe(py)) or 2(Fe-Fe(py)₂), and these species could

form different types of aggregates due to π-π stacking

interactions. The mixture of multiple coordination types

and aggregation states gives a very broad Q-band at 684

nm and a shoulder (589 nm) (stage B in Fig. 3b). The low

energy peak at 745 nm is reduced in intensity, but the Soret

band at 335 nm was still depressed due to aggregation.

At very high pyridine concentrations, tri- or tetra-

pyridyl-2 molecules with spectroscopic characteristics of

isolated pc dimers are observed. When neat pyridine was

used as the solvent, the Q-band and Soret band reached

the highest intensity (phase C in Fig. 3b). The signatures

for disaggregation in phase C include: an intense Q-band

(696 nm, log ε = 4.57), loss of the low-energy band at

745 nm, and a strong Soret band (344 nm, log ε = 4.70).

Several sharp vibronic shoulder peaks were also visible

(651 nm, 634 nm and 589 nm). The ligand-metal charge

transfer peak was not visible. It could be blue-shifted and

buried under the peak at 420 nm. The peak at 420 nm was

assigned as the oxygen-to-π* charge-transfer band of 2.

This assignment is supported by the fact that (Figs S8a,b)

the peaks at 420 nm remain at the same position regard-

less of the change of the concentration of pyridine.

As shown in Table 1, the chemical shift of the phenyl

proton of 1 (H_a) appears as a slightly broadened singlet at

8.66 ppm, a value which is typical for aromatic protons on

a phthalocyanine core. The chemical shift of the “bridg-

ing” phenyl proton of 2 (H_g) is 9.74 ppm, which is rea-

sonable considering the inﬂuence of the two 18-member

aromatic rings. The other phenyl protons in compound 2

(H_h, H_gand H_i) had chemical shift values of 8.72 ppm,

8.69 ppm and 8.68 ppm, respectively. Strong residual

pyridine peaks at (8.50, 7.55, 7.12 ppm) are visible in

nMr

Previous studies on phthalocyanine monomer and

dimer molecules have had great difﬁculties to obtain well

Fig. 5. NMR spectra of 1 in CDCl₃with py-d₅. (a) full ﬁeld (b) low ﬁeld region

J. Porphyrins Phthalocyanines 2011; 15: 283–292

284

W. He anD M. LIeberMan

Fig. 6. NMR spectra of 2 in CDCl₃with py-d₅

table 1. Proton NMR assignment of 1 and 2^a

H_aH_bH_cH_d,H_e

ppm 8.66 (4) 4.42 (8) 2.17 (8) 1.50 (16)

H_gH_hH_iH_j

starting value. The current at the working electrode is

recorded at end of every cycle and plotted against the

potential change. The reduction or oxidation peak can

appear as a peak or trough, depending on the sweep

direction. All the experiments were kept at initial poten-

tial for several seconds before starting to scan, and all

the SW voltammograms were taken with 5 sec of quiet

time.

1

H_f

0.92 (12)

H_k

2

ppm 9.74 (3^b) 8.72 (4) 8.69 (4) 8.68 (4) 4.55-4.37 (24)

H_lH_mH_nH_o

Since the electrochemistry data of 1 and 2 are com-

plicated and interaction with pyridine also shifts their

redox potentials, the data will be presented and dis-

cussed in three subsections. First, the cyclic voltam-

metry and square-wave voltammetry data of 1 and 2 are

presented. Second, the literature data of pc ring and cen-

tral metal redox behaviors are summarized, along with

discussion of the inﬂuence of solvent and ligands on the

redox potentials of metallophthalocyanines. Finally, the

assignment of the redox waves in compounds 1 and 2 is

discussed.

Electrochemistry results for 1 and 1-py. Electro-

chemical studies were done in methylene chloride. With-

out adding pyridine to 1, the CV and SW spectra were not

very clear or interpretable. Some typical CV and SW vol-

tammograms of 1 are shown in the Supplementary mate-

rial (Figs S11–S13). In order to break up aggregation of

the phthalocyanine rings, 30 mM pyridine was added to

1.0 mM 1; under these conditions, the main species is

the cofacially aggregated dimer of 1-py. Figure 7 shows

the resulting SW voltammogram. The cathodic sweep

showed peaks at 1.29, 0.71, -0.90, and -1.21 V, and the

anodic sweep showed peaks at 0.86, 0.68, 0.19, -0.41,

-0.75, and -0.88 V.

2.12 (24) 1.70 (24) 1.53 (24) 1.03 (36)

a

The numbers in brackets are the integrated intensity values.

^bThe integration of Hg is discussed in the text.

the NMR of 1 with py-d₅(Fig. 5b) and at (8.59, 7.64,

7.27 ppm) in the NMR of 2 with py-d₅(Fig. 6b); they

were identiﬁed by comparison to the reference data of

pyridine in chloroform (8.62, 7.68, 7.29 ppm) [74]. No

peaks were observed for coordinated pyridine because

of the low quantity of protonated pyridine present in

the samples. For compound 2, some of the integrations

could not be measured well. For example, the integration

of H_gwas 50% more than the theoretical value due to a

small overlapping impurity peak. Several small peaks

are clearly present in the aromatic region, which may

indicate the presence of multiple coordination states or

aggregation states in the sample. The CH₂O protons of

the pentyloxy side chains appeared as clear triplets at

4.42 ppm and 4.46 ppm for 1 and 2, respectively, and

the peak positions and intensities for the rest of the

pentyloxy chains were consistent with previous reports

[16, 64, 75].

Known electrochemistry of phthalocyanine mono-

mers. For Fe(II)Pc, the ﬁrst ring oxidation occurs at

potentials from 0.74 to 1.22 V, depending on the solvent

and whether there is a coordinating ligand present; the

ﬁrst ring reduction occurs at potentials ranging from -1.1

to -1.5 V and is also affected by the presence of coor-

dinating ligands [76, 77]. As seen in Table 2, oxidation

of the central Fe(II) in FePc is observed at potentials

between 0.17 and 0.66 V, with more positive oxidation

potentials observed in the presence of π-backbonding

ligands such as pyridine. Reduction of Fe(II) in Fe(Pc)

is observed at potentials around -0.86 to -1.07 V depend-

ing on the chelation ability of solvents and eletrolytes.

Electrochemistry

Electrochemical data were obtained with cyclic vol-

tammetry (CV) and square-wave voltammetry (SW) in

a nitrogen-ﬁlled dry box with electrical feedthroughs.

The working electrode was a Pt disk, and we found it

necessary to polish the electrode before each run. In the

cyclic voltammetry experiment, the applied potential is

systematically swept over a desired range while measur-

ing current at the working electrode. In the square-wave

voltammetry experiment, the voltage is held at a starting

value and a square-wave pattern is used to jump up to

sequentially higher voltages, each time returning to the

J. Porphyrins Phthalocyanines 2011; 15: 284–292

THe SynTHeSIS anD CHaraCTerIzaTIOn Of a SIDe-by-SIDe IrOn PHTHaLOCyanIne DIMer

285

Fig. 7. Square-wave voltammograms of 1 in CH₂Cl₂with 1.00 mM 1, 30.0 mM pyridine (a: anodic scan and b: cathodic scan)

table 2. Redox potentials of FePc complexes^a

assignment was conﬁrmed by electron spin resonance

(ESR) data. The 0.71 V process was electrochemically

irreversible (no return wave in CV). The last oxidation

wave at 1.29 V is assigned as the ﬁrst ring oxidation

of the pc ring in 1; this ring oxidation was observed at

1.05–1.10 V in the parent compound FePc in pyridine

solvent [78].

Solvent electrolyte Fe^III/Fe^IIFe^II/Fe^IFirst ring reduction

pyridine

DMA

TEAP^b

TEABr^c

TEAP^b

TEABr^c

0.66

-1.07

-1.05

-0.55

-0.86

-1.316

-1.283

-1.169

-1.194

0.38

0.17

When Fe^IIIPc-py is formed, the coordinated pyridine

falls off [80]. The resulting mixture of py-coordinated 1

and uncoordinated 1 gives the complex return wave in

Fig. 7b, which shows peaks at 0.86, 0.68, 0.19, -0.41,

-0.75, and -0.88 V.

a

b

0.0005 M FePc, V vs. SCE, [80]. 0.1 M solution of TBAP

(tetrabutylammonium perchlorate). ^c0.1 M solution of TEABr

(tetrabutylammonium bromide).

There are few reports of Fe-centered redox processes in

non-coordinating solvents and in the absence of coordi-

nating ligands. Wolberg and Manasson [78] reported the

Fe^III/Fe^IIredox potential for FePc in chloronapthalene

at 0.19 V, and Campbell et al. [79] report a reduction

potential of -0.95V for the Fe^II/Fe^Iprocess for Fe[t-BuPc]

in CH₂Cl₂(DCM).

Discussion of electrochemistry of 1 and 1-py. When

compound 1 was studied by cyclic voltammetry or square-

wave voltammetry, only weak or indistinct peaks were

observed, even at mM concentrations. This is probably

due to the aggregation of the pc molecules. The clear-

est voltammograms were measured when excess pyridine

was added to the solution to form pyridine-coordinated

1. According to the UV-vis titration results, for a 1 mM

solution of 1 in the presence of 30 equivalents of pyridine

the major species is the 1:1 complex 1-py, which would

mostly exist as a dimer.

The initial anodic SW sweep is shown in Fig. 7a. Peaks

were observed at -1.21, -0.90, +0.71, and +1.29 V. The

reduction wave at -1.21 V is assigned as the ﬁrst reduc-

tion of the pc ring in 1; this reduction occur at -1.27 V

in the parent compound FePc (in pyridine solvent). The

reduction at -0.90 V is assigned as the reduction of the

Fe(II) in 1; in the parent compound FePc this reduc-

tion was observed at -1.07 V in the presence of pyridine

(the Fe reduction potential in FePc shifts to -0.55 V in

a non-coordinating solvent, DMA). The ﬁrst oxidation

peak (0.71 V) in Fig. 7a was assigned to the oxidation

of Fe^IIin 1-py, where the Fe(II) is stabilized by pyridine

coordination. In prior work by Lever [80], the ﬁrst oxida-

tion potential of FePc occurred at 0.66 V in neat py; the

The oxidation peak at 0.86 V was assigned to the ﬁrst

ring oxidation of py-free 1. The peak at 0.68V is assigned

to the oxidation of the Fe(II) in 1-py (seen at 0.71 V in the

cathodic sweep). The weak peak at 0.19 V in Fig. 7a is

assigned to oxidation of the Fe(II) in py-free 1, which is

the same as the potential for oxidation of Fe(II) in FePc

in chloronapthalene, 0.19 V [78]. Comparing the Fe^III/

Fe^IIreduction potentials of py-coordinated (0.71 V) and

py-free (0.19 V) 1, we could see that Fe^IIPc was harder

to oxidize in py-containing solution, presumably due to

the back-bonding interactions between the Fe(II) cen-

ter and the pyridine pi system. The reduction wave at

-0.41 V is assigned as the ﬁrst reduction of Fe^IIin py-free

1. In the voltammograms of 1 (Fig. 7a,b), the separation

between the ﬁrst Fe oxidation and ﬁrst Fe reduction was

smaller for py-free 1 (oxidation 0.19, reduction -0.41,

separation of 0.60V) than for py-coordinated 1 (oxida-

tion 0.71V, reduction -0.90 V, separation of 1.61 V). The

reported separation [80] between Fe(II) oxidation and

reduction potentials in FePc ranges from 0.5–1.2 V for

weakly coordinating solvents but rises to 1.7–1.9 V in

neat pyridine. The second reduction peak for compound

1 at -0.75 V is tentatively assigned as the ﬁrst ring reduc-

tion of py-free 1. The ﬁnal reduction peak (-0.88 V) was

assigned as the reduction of the Fe(II) in 1-py, seen at

-0.90 V in the cathodic sweep. Thus the pyridine-free 1 is

associated with redox peaks at 0.86 V, 0.19 V, -0.40 V and

-0.75 V. In cyclic voltammograms of py-free 1, Fig. S12

(Supplementary material), although the capacitive back-

ground was high, weak peaks at 0.763 V, 0.145 V, -0.50 V

and -0.84 V were identiﬁed. All the peak assignments are

listed in Table 3.

J. Porphyrins Phthalocyanines 2011; 15: 285–292

286

W. He anD M. LIeberMan

a

table 3. Electrochemical data for 1 (E_1/2/V )

reductions form two pairs with separations of about

250 mV each. The ﬁrst ring reduction couple is shifted

positively by about 300 mV relative to the value for the

analogous monomeric phthalocyanine, and the ﬁrst ring

oxidation couple is shifted negatively by about 70 mV.

The CV and DPV of (CoPc)₂were also obtained [33].

Because of cobalt’s redox activity, the voltammograms

were more complex than for (ZnPc)₂. In dichloroben-

zene solution, the splitting of the two successive Co^II/Co^I

couples was fairly large (about 500 mV), and the ratio of

area of the peaks deviated from the theoretical ratio. The

authors assigned four redox couples to stepwise reduc-

tions of the two phthalocyanine units in (CoPc)₂. The

redox potentials for the ﬁrst metal oxidation in CoPc and

[CoPc]₂are virtually identical.

Couple

E_1/2/V

1-py

Fe^IIIPc(-1)/Fe^IIIPc(-2)

Fe^IIIPc(-2)/Fe^IIPc(-2)

Fe^IIPc(-2)/Fe^IPc(-2)

Fe^IPc(-2)/Fe^IPc(-3)

Fe^IIIPc(-1)/Fe^IIIPc(-2)

Fe^IIIPc(-2)/Fe^IIPc(-2)

Fe^IIPc(-2)/Fe^IPc(-2)

Fe^IPc(-2)/Fe^IPc(-3)

1.29

0.71

-0.90

-1.21

0.86

1

0.19

-0.40

-0.75

^aPotentials are reported with respect to the SCE couple.

Electrochemistry results for 2 and 2-py. In the absence

of pyridine, the CV and SW spectra of 2 were not very

clear or interpretable. Some typical CV and SW voltam-

mograms of 2 are shown in the Supplementary material.

Pyridine was added to 2 to break up the aggregation.

Figure 8 shows the SW voltammogram of 1.2 mM 2 in

the presence of 857 mM pyridine. The spectrum shows

peaks assigned as py-coordinated 2 (see Discussion of

electrochemistry of 2 and 2-py section hereafter) in the

cathodic (1.45, 1.20, 0.84, -0.67, -0.89, -1.23, -1.28 V)

and anodic (1.16, 0.93, 0.52, -0.43, -0.65, -0.90, -1.21 V)

scans.

As the inﬂuence of pyridine on the redox potentials

of iron phthalocyanine dimers has not been established,

we used the data of the inﬂuence of pyridine on 1 to help

assign our electrochemistry data for 2.

Discussion of electrochemistry of 2 and 2-py. The

cyclic voltammograms and square wave voltammograms

of pure 2 were weak and complicated, but after addition

of about 700 equivalents of pyridine, the voltammograms

became clearer. According to our UV-vis experiment, at

these concentrations, 2 was about halfway occupied by

pyridine to form 2-(py)₂, which still could aggregate to

form cofacial dimers.

Known electrochemistry of phthalocyanine dimers.

As seen in Table 4, Kobayashi observed six redox pro-

cesses for a binuclear pc (with six solubilizing neopenty-

loxy side chains) whose aromatic core is nearly identical

to that of 2 [33]. This compound, which lacks a redox

active metal, showed two quasi-reversible ring oxidations

at 0.66 and 0.41 V vs. SCE. and four quasi-reversible ring

reductions (-0.85, -1.11, -1.45, and -1.70 V vs. SCE) in

the non-coordinating solvent dichlorobenzene. The ring

oxidations are separated by about 250 mV, and the ring

In the cathodic square-wave scan, three oxidation

waves and four reduction waves were observed (at 1.45,

1.20, 0.84, -0.67, -0.89, -1.23, and -1.28V vs. SCE,

Fig. 8a and Table 5). The anodic scan showed most of

the expected return waves along with two new waves

(at 1.16, 0.93, 0.52 (new), -0.43 (new), -0.65, -0.90, and

-1.21 V, Fig. 8b and Table 5). We think the new peaks

arise due to loss of pyridine coordinated to 2 during the

oxidative preparation, which allows peaks that belong to

redox processes of py-free 2 to appear. We assumed that

Fig. 8. Square-wave voltammograms of 2 in CH₂Cl₂with 1.20 mM 2, 857 mM pyridine (a: anodic scan and b: cathodic scan)

table 4. Redox potentials of side-by-side pc dimer (ZnPc)₂complexes^a

Second ring

oxidation

First ring

oxidation

First ring

reduction

Second ring

reduction

Third ring

reduction

Fourth ring

reduction

0.66

0.41

-0.85

-1.11

-1.45

-1.70

^a0.002 M pc dimer in 0.1–0.3 M solution of TBAP in dichlorobenzene, in V vs. SCE [33].

J. Porphyrins Phthalocyanines 2011; 15: 286–292

THe SynTHeSIS anD CHaraCTerIzaTIOn Of a SIDe-by-SIDe IrOn PHTHaLOCyanIne DIMer

287

a

table 5. Electrochemical data for 2 (E_1/2/V )

reduction than py-free FePc [80], the peak at -0.43 V was

assigned to the ﬁrst Fe^II/Fe^Ireduction of py-free 2.

Couple

E_1/2/V

Similarly, the peak at -0.89 V in the py-coordinated 2

was about twice the size of other peaks for py-coordinated

2. This peak was assigned to the overlapping couples

of [Fe^IIPc(-2)Fe^IPc(-2)]/[Fe^IPc(-2)]₂and [Fe^IPc(-2)]₂/

[Fe^IPc(-3)Fe^IPc(-2)] of the py-coordinated 2. The reduc-

tion potentials of the two steps were expected to be very

close and we could not tell the sequence of these two steps.

Since the peak at -0.65 V of py-free 2 was also twice the

size of the other peaks, it was treated as a two electron addi-

tion process, composed of adding one electron to the metal

and the other electron to the ring. Therefore, the -0.65 V

was assigned to the [Fe^IIPc(-2)Fe^IPc(-2)]/[Fe^IPc(-2)] and

[Fe^IPc(-2)]₂/[Fe^IPc(-3)Fe^IPc(-2)] of py-free 2.

The peaks at -1.23 V and -1.28 V were assigned as the

consecutive reductions on the pc rings of py-coordinated 2

since it was not enough to reduce Fe^Ito Fe⁰. Similarly, the

-0.99V and -1.21V were assigned to the consecutive reduc-

tions on the pc rings of py-free 2. The trough at -0.90 V was

much bigger than the others and we are not sure of the cause.

In Kobayashi’s paper [33], the sizes of the peaks were also

not proportional the number of the electron exchanges.

As discussed above, both 1-(py) and 2-(py)₂could

lose their axial pyridine ligands if the Fe^IIwas oxidized

to Fe^III. Figure 7b shows redox waves of both py-free and

py-coordinated 1 molecules, which conﬁrmed partial

dissociation of py-FePc during the 5 sec of electrolysis

at 1.5 V. The same process happens for py-coordinated

2 during the 5 sec of cathodic electrolysis. In Fig. S13

(Supplementary material), peaks of py-free 2 at 0.62 V

and 0.92 V were close to the ﬁrst two oxidation peaks

(0.52 V, 0.92 V) in the cathodic scan of py-coordinated 2

(Fig. 8b). The ﬁrst reduction peak of py-free 2 (-0.64 V)

was not very close to that (-0.43 V) of the cathodic scan

of py-coordinated 2 (Fig. 8b). The difference between the

cathodic voltammograms of py-coordinated 1 and py-co-

ordinated 2 was that in Fig. 7b a mixture of py-free and

py-coordinated 1 can be seen, while in Fig. 8b, py-free 2

appears to be the major species present.

2-py₂

[Fe^IIPc(-1)Fe^IIIPc(-1)]/[Fe^IIPc(-1)Fe^IIIPc(-2)]

[Fe^IIPc(-1)Fe^IIIPc(-2)]/[Fe^IIPc(-2)Fe^IIIPc(-2)]

[Fe^IIPc(-2)Fe^IIIPc(-2)]/[Fe^IIPc(-2)]₂

[Fe^IIPc(-2)]₂/[Fe^IIPc(-2)Fe^IPc(-2)]

[Fe^IIPc(-2)Fe^IPc(-2)]/[Fe^IPc(-2)]₂

[Fe^IPc(-2)]₂/[Fe^IPc(-3)Fe^IPc(-2)]

1.45

1.20

0.84

-0.62

-0.89

-1.23

-1.28

1.46

[Fe^IPc(-3)Fe^IPc(-2)]/[Fe^IPc(-3)]₂

[Fe^IPc(-3)]₂/[Fe^IPc(-3)Fe^IPc(-4)]

2

[Fe^IIIPc(-1)]₂/[Fe^IIPc(-1)Fe^IIIPc(-1)]

[Fe^IIPc(-1)Fe^IIIPc(-1)]/[Fe^IIPc(-1)Fe^IIIPc(-2)]

[Fe^IIPc(-1)Fe^IIIPc(-2)]/[Fe^IIPc(-2)Fe^IIIPc(-2)]

[Fe^IIPc(-2)Fe^IIIPc(-2)]/[Fe^IIPc(-2)]₂

[Fe^IIPc(-2)]₂/[Fe^IIPc(-2)Fe^IPc(-2)]

[Fe^IIPc(-2)Fe^IPc(-2)]/[Fe^IPc(-2)]₂

[Fe^IPc(-2)]₂/[Fe^IPc(-3)Fe^IPc(-2)]

1.16

0.93

0.52

-0.43

-0.65

-0.99

-1.21

[Fe^IPc(-3)Fe^IPc(-2)]/[Fe^IPc(-3)]₂

[Fe^IPc(-3)]₂/[Fe^IPc(-3)Fe^IPc(-4)]

^aPotentials are reported with respect to the SCE couple.

the shifts in potential between FePc and (FePc)₂will be

similar to the corresponding shifts in potential for known

CoPc and (CoPc)₂redox transitions and that the inﬂu-

ence of pyridine on the electrochemistry of (FePc)₂will

be similar to that of FePc.

Since we assigned 0.71V to py-coordinated Fe^IIIPc(-2)/

Fe^IIPc(-2) in 1 and 0.19 V to the same transition in py-

free 1, the anodic 0.84 V wave (observed at 0.93 V in the

cathodic scan) of compound 2 probably belongs to the

ﬁrst Fe^IIoxidation ([Fe^IIPc(-2)Fe^IIIPc(-2)]/[Fe^IIPc(-2)]₂)

of py-coordinated 2 and the new peak at 0.52 V to the

ﬁrst Fe^IIoxidation of py-free 2.

For compound 1-py, the ﬁrst ring oxidation was

observed at 1.29 V; a shift of 70 mV negatively should

give a peak around 1.22 V. There are two possible assign-

ments, but since the second ring oxidation should be

about 250 mV more positive than the ﬁrst ring oxidation,

the peaks at 1.20 and 1.45 V was assigned to the ﬁrst

and second pc ring oxidation of py-coordinated 2. Like-

wise, py-free 1 showed its ﬁrst ring oxidation at 0.86 V;

a 70 mV negatively should give a peak around 0.79 V.

The oxidation peak at 1.46 V for py-free 2 is assigned to

the second Fe^III/Fe^IIoxidation. The peaks at 0.93 V and

1.16 V were accordingly assigned to the ﬁrst and second

ring oxidations of py-free 2.

Mixed-valence properties of the side-by-side iron(pc)

dimer 2. Electrochemical data can give information

about the stability of mixed-valence compounds which is

a key metric for use in molecular QCA. Scheme 5 shows

two potential formation reactions, oxidation I and reduc-

tion II, for mixed-valence compounds based on 2. Using

Equation III, where the ∆E is the difference in redox

potentials for the two compounds shown in each forma-

tion reaction, we can calculate comproportionation con-

stants (K_cat 20 °C) listed in Table 6. Here the two solvent

From the results of Lever [80], the separation between

the ﬁrst Fe^IIoxidation and reduction of a py-coordinated

FePc monomer was about 1.73 V. This observation made

us chose -0.65 V as the ﬁrst Fe^IIreduction for py-coordi-

nated 2. Because the py-coordinated FePc monomer had

a larger separation between the ﬁrst Fe^IIoxidation and

Scheme 5. Mixed-valence formation reactions of 2

J. Porphyrins Phthalocyanines 2011; 15: 287–292

288

W. He anD M. LIeberMan

table 6. Comproportionation data

e

Couple^a

Solvent^b

∆E/V^c

K_c^d

K_d

∆G(Kj/mol)

I

A

B

A

B

0.41

0.36

0.22

0.26

1.1 × 10⁷

1.5 × 10⁶

6.1 × 10³

3.0 × 10⁴

9.1 × 10^-8

6.4 × 10^-7

1.7 × 10^-4

3.4 × 10^-5

-39

-35

-21

-25

I

II

^aI, II correspond to couples I, II in Scheme 5. ^bA = CH₂Cl₂, B = 0.86 mM pyridine in CH₂Cl₂. ^cMixed-

valence splitting energy. ^dComproportionation constant. ^eDiscomproportionation constant, 1/K_c.

environments were appointed as CH₂Cl₂(environment A)

and 0.86 mM pyridine in CH₂Cl₂(environment B). In the

CH₂Cl₂environment the py-free 2 values were used and

in the pyridine solution, 2-py₂data were used.

and chloroform were dried by storage over activated

3 Å molecular sieves. KBr for IR spectra was kept in a

desiccator when not in use.

The comproportionation constant (K_c) of oxidation

reaction I was over 1.6 × 10⁶in both solvents – 1.1 ×

10⁷in pure DCM, and 1.6 × 10⁶in DCM with 0.86 mM

pyridine. The difference is probably not signiﬁcant given

the difﬁculty of pinpointing the redox potentials. The

K_comfor 2⁺is comparable to the value for a well known,

stable mixed-valence compound, the Creutz-Taube ion,

([(NH₃)₅Ru-pz-Ru(NH₃)₅]⁵⁺, for which K_com= 4 × 10⁶)

[19]. The comproportionation constant (K_c) for 2^-(formed

by reduction reaction II) was smaller, measuring 6.1 ×

10³(DCM) and 3.0 × 10⁴(DCM/py).

Multiple attempts were made to electrolyze 2 in the

presence of pyridine to form the oxidized mixed-valence

species. Several attempts failed to give identiﬁable prod-

ucts. Next chemical oxidation of the monomer 1 was

tried. Following the experimental procedure of Song

[81], one equivalent of CF₃COOAg (1.06 V vs. AgCl/Ag)

was mixed with 1 in CH₂Cl₂for 10 min in the dark. The

product was brown in color and the UV-vis spectra was

not consistent with other known [Fe^IIIPc]⁺species [22].

We think the pc ring was oxidized. We attempted to get

EPR data at room temperature but only Ag⁺was visible

in the EPR spectrum.

Spectroscopy

¹H NMR spectra were taken on Varian UnityPlus

300 or AVANCE Bruker DPX400 or Varian INOVA500

instruments and were referenced to the appropriate

solvent residual peaks. UV-vis spectra were taken on a

Perkin-Elmer Lamba 11 UV-vis spectrometer. IR spectra

were taken as KBr pellets or as thin ﬁlms from evapo-

ration of chloroform or ethanol on KBr salt plates and

were measured on a Perkin-Elmer Paragon 1000 FTIR

spectrometer.

Mass spectra were taken on a JEOL JUSAX505 HA

mass spectrometer (ESI, FAB) or on a Voyager Perspec-

tive Biosystem mass spectrometer (MALDI). For ESI,

the sample was dissolved in an electrospray solvent such

as 50:50 v:v methanol:water with 2% acetic acid. This

solution was passed through a metal capillary which

was biased at high potential (4–5 kV). For MALDI, a

2,5-dihydroxybenzoic acid (DHBA) or α-cyano-4-

hydroxycinnamic acid (ALPHA) MALDI matrix was

dissolved in water or ethanol at a concentration of 20 mg/

mL. The chloroform solution of 1 or 2 was mixed with

the matrix in a ratio of 1:100 v/v in a 0.5 mL Eppendor-

f^TMmicrotube. 1µL of this mixture was deposited on the

gold sample plate, dried at room temperature and then

ionized by laser and analyzed. A nitrogen 32 UV Laser

operating at 337 nm was used as the ionization source.

Since chloroform and water are not miscible, the phthalo-

cyanine signal was not evenly spread on the sample plate

and we needed to move the ionization location around to

maximize the signal.

ExPErIMEntAl

General

The following materials were commercially available,

and were used without puriﬁcation: bromine, catechol,

copper(I) cyanide, potassium carbonate, magnesium

sulfate, DMF, toluene, THF, pyridine-d₅, 1,2,4,5-tetracy-

anobenzene and ferrous dichloride tetrahydrate. Pyridine

was dried over calcium hydride, then vacuum transferred

and stored in a dry box under prepuriﬁed nitrogen. N,N

dimethylaminoethanol and 2-chloronaphthalene were

freshly distilled whenever needed. Toluene, benzene

and benzene-d₆were dried over sodium, vacuum trans-

ferred and stored in a dry box under prepuriﬁed nitro-

gen. Ethanol, methanol, acetone, dichloromethane,

HPlc

HPLC puriﬁcation was done using a Waters^TMalliance

2695 separation module along with a Waters^TM996 pho-

todiode array detector. A Phenomenex^TMLUNA silica

column (250 × 4.6 mm, 100 Å) was used as the analyti-

cal HPLC column. A Phenomenex^TMLUNA silica semi-

prep column (250 × 10 mm, 100 Å) was used for large

scale separation. 0.5 or 1mL/min ﬂow rates were used

for analytical and semi-prep silica columns, respectively.

J. Porphyrins Phthalocyanines 2011; 15: 288–292

THe SynTHeSIS anD CHaraCTerIzaTIOn Of a SIDe-by-SIDe IrOn PHTHaLOCyanIne DIMer

289

The eluent was an isocratic mixture of 2% pyridine in

chloroform. Care should be taken to vent the eluent and

its wastes to a fume hood.

0.5 mol) was dissolved in 200 mL acetic acid and kept in

an ice bath, and bromine (50 mL, 0.97 mol) was added

via a dropping funnel over an hour. After the bromine

was added, the reaction mixture was stirred for another

2 h on ice. After the solvent was removed by rotary

evaporation, (CAUTION: HBr and HOAc are caustic;

vacuum should be provided by an arrangement such as

a diaphragm pump inside the ventilation hood to prevent

HBr release to the lab) the brown oil was poured into

1.5 L ice-water. The precipitate was ﬁltered, dried in air

and recrystallized from toluene to give pale white crys-

Electrochemistry

Cyclovoltametric measurements were performed in

a nitrogen-ﬁlled dry box using an EG&G instruments

PAR 283 potentiostat/Galvanostat or BAS Epsilon EC

instrument. A conventional three-electrode system was

used. The working electrode was a 0.02 cm²Pt disk. An

Ag/AgCl wire (soaked in 0.1 M tetrabutylammonium

chloride(TBACl)/dichloromethane solution, 0.29 V vs.

NHE or 0.044 V vs. SCE) was used as the reference elec-

trode. All redox potentials were converted to the values

vs. SCE. A platinum mesh spot welded to a platinum

wire was used as the counter electrode. HPLC puriﬁed

1 (> 99%) and 2 (> 98%) were used for electrochemis-

try, typically at concentrations of 1 mM. Tetrabutylam-

monium tetraﬂuoroborate (TBABF₄, 0.1M in CH₂Cl₂)

served as the electrolyte. When pyridine was added, it

was present in large excess (10 mM–1M). Scan rates

from 50 mV/s to 200 mV/s were used for cyclovoltam-

metry. In square wave voltammetry, the square wave

amplitude and frequency were set to 50 mV and 50 Hz,

respectively. The current full scale was 1 µA and a 1.0 Hz

ﬁlter was used. The quiet time was set to 5 s and 1 mil-

lisecond was used as the sample period. For controlled-

potential electrolysis of 2, the potential was set to 1.5 V

in order to get complete oxidation.

1

tals (52 g, 45% yield). H NMR (CDCl₃): δ, ppm 7.16

(2H, s, Ar-H), 5.72 (2H, s, OH).

1,2-dibromo-4,5-bis(pentyloxy)benzene (5). The

synthesis procedure of this step followed Hanack et al.

[36]. Dibromocatechol (4) (26.8 g, 0.10 mol) and potas-

sium carbonate (20 g, 0.14 mol) were dissolved in 50 mL

DMF, stirred for 30 min, then 1-bromopentane (25 mL,

0.21 mol), dissolved in 20mL DMF, was added dropwise.

The resulting brown solution was heated to 110 °C for

12 h, then the dark brown reaction solution was poured

into 500 mL 0.2 N HCl and the product was extracted

with 40 mL ether three times. The ether was removed by

rotary evaporation, yielding 15 mL of brown oil. After

recrystallization from ethanol, 34.2g white crystals were

1

obtained (85% yield). H NMR (CDCl₃): δ, ppm 7.07

(2H, s), 3.95 (4H, t, J = 6.6 Hz), 1.82 (4H, m), 1.40 (8H,

m), 0.92 (6H, t, J = 6.9 Hz). IR (thin ﬁlm from ethanol):

ν, cm^-12957 (vs), 2872 (vs), 1583 (m), 1498 (vs), 1468

(vs), 1387 (m), 1352 (s), 1329 (m), 1251 (vs), 1201 (vs),

1116 (m), 989 (m), 880 (m), 652 (m).

The cleanliness of the working electrode is very

important to get good SW voltammograms. We used Pt

disk electrodes consisting of a highly polished Pt wire

embedded in a solvent-resistant chlorotriﬂuoroethylene

(CTFE) plastic body. The small diameter of the wire

(1.6 mm) gave a higher intensity of current than bare Pt

or Au wires. It was important to polish the contacting end

(diamond paste) before each experiment.

1,2-dicyano-4,5-bis(pentyloxy)benzene (6). This

synthesis followed van Nostrum’s procedure [82]. CuCN

(1.8 g, 150 mmol) was added to a solution of 5 (20.4 g, 50

mmol) in 120 mL DMF, and the solution was heated to

150 °C for 12 h under Ar. The green mixture was poured

into 600 mL concentrated aqueous NH₄OH after it cooled

down to RT, and air was bubbled through for 6 h. The

green precipitate was ﬁltered off and washed with water

until the ﬁltrate was neutral. After the solid air dried, it

was extracted into 200 mL methanol in a Soxhlet extrac-

tor for 24 h. The product was crystallized from methanol

(50 mL). It was puriﬁed by recrystallization in ethanol

(30 mL) (white ﬂaky crystals, 8.9 g, yield 59%). ¹H NMR

(CDCl₃): δ, ppm 7.12 (2H, s), 4.05 (4H, t, J = 6.6 Hz),

1.86 (4H, m), 1.44 (8H, m), 0.94 (6H, t, J = 7.0 Hz). IR

(thin ﬁlm from CH₂Cl₂): ν, cm^-12228 (s, CN), 1590 (vs),

1526 (s), 1466 (m), 1394 (m), 1365 (m), 1293 (m), 1232

(vs), 1218 (s), 1095 (vs), 975 (m), 888 (m). MS (FAB):

m/z calcd. 301 (100%) [M + H]⁺, 302 (20%) found 301

(100%) [M + H]⁺, 302 (25%), 275 [M + HCN]⁺, 249

[M + H₂CN]⁺, 162 [M + H₂CN–OC₅H₁₁]⁺.

uV-vis titrations with pyridine

Samples of pure 1 and 2 were dried under vacuum on

a Schlenk line for 2 h at room temperature at 20 mTorr

to remove the pyridine residue from HPLC puriﬁcation.

Then, 1 (0.16 mg) or 2 (0.13 mg) was dissolved in 2.00

mL CHCl₃in a UV-vis cuvette. The concentrations of 1

and 2 were 6.4 × 10^-5M and 3.1 × 10^-5M, respectively.

Pyridine was added with a 10 µL glass syringe. Four

kinds of pyridine solutions were made for injection: 0.1%

(v/v) py/CHCl₃, 1% (v/v) py/CHCl₃, 10% (v/v) py/CHCl₃

and neat pyridine. At the end of each titration, a total of

130 µL solution had been added to the cuvette. Since the

volume change was only 6%, we considered the concen-

tration of the pc to be constant during the process.

Bisdiiminoisoindoline (8). This process was modiﬁed

from Piechocki’s thesis [83]. The bis(diiminoisoindole)

(8) was prepared by bubbling gaseous ammonia through

1,2,4,5-tetracyanobenzene (1.0 g) (7) in ethanol solu-

tion for 5 h. Since 8 has very low solubility in methanol,

Synthesis

1,2-dibromocatechol (4). The synthesis procedure of

this step followed Kohn’s work [37]. Catechol (3) (55 g,

J. Porphyrins Phthalocyanines 2011; 15: 289–292

290

W. He anD M. LIeberMan

it precipitated as a yellow powder (1.2 g, 100%) and

was isolated by ﬁltration. The reaction is conveniently

followed by IR, as the peak at 2245 cm^-1(C≡N) of the

starting material disappears. EI: m/z calcd. 212 (100%)

[M + H]⁺, 213 (13%), found 212 (100%) [M + H]⁺, 213

(30%), 198 [M + HN]⁺. IR (KBr pellet): ν, cm^-13359

(vs), 2950 (vs), 1686 (vs), 1560 (m), 1402 (vs), 1070 (m),

960 (s), 833 (s), 544 (s).

Peaks eluted at 26.5 and 30 min, which corresponded

to compounds 1 and 2, respectively. 2 was checked by

analytical HPLC to conﬁrm its purity (> 98% measured

at 699 nm). The yield of pure 2 (38 mg) was 12% based

on tetracyanobenzene. Even after prolonged drying

under 20 mTorr vacuum on a Schlenk line, pyridine

from the HPLC separation seems to be present in both

1 and 2 (as seen in the NMR) and we cannot rule out the

presence of other impurities that are NMR and HPLC

silent. The UV spectrum was measured in neat pyridine

solution to get characteristic isolated phthalocyanine

Side-by-side iron phthalocyanine dimer (2) and

iron phthalocyanine monomer (1). The synthesis of

these compounds followed the statistical method of

Bossard’s work on ruthenium phthalocyanines [84]. A

50 mL three neck ﬂask was ﬁlled with bisdiiminoisoin-

doline (0.038 g, 0.18 mmol) (8) and 10 mL of DMAE

and heated to boiling temperature under nitrogen. The

reaction was set to reﬂux through one neck under a

slow nitrogen purge (to a gas bubbler). The other two

necks were sealed with SUBASEAL^TMrubber stoppers.

Ferrous dichloride tetrahydrate (FeCl₂·4H₂O, 0.066 g,

0.33 mmol) was boiled in 5 mL of DMAE for 5 min

until all the water had distilled out. The iron solution

was added slowly over 20 min through one neck of the

ﬂask using a syringe and the syringe was rinsed with a

little DMAE to ensure quantitative transfer. Simultane-

ously, through the other neck, excess (0.30 g, 1.00 mmol)

1,2-dicyano-4,5-bis(pentyloxy)benzene (6) in 5 mL of

DMAE was added in 20 min using another syringe.

After all the ingredients were added, the syringes were

removed, the nitrogen was turned off, and a long needle

was connected to an ammonia gas tank via a gas trap (to

prevent any suckback of the reaction mixture into the

tank) to introduce ammonia into the reaction ﬂask. The

resulting brown suspension was gently reﬂuxed under a

slow bubbling of ammonia gas for 3 days. After cooling

to room temperature, 100 mL of ethanol was added to

precipitate the phthalocyanine products. After washing in

a Soxhlet extractor with acetone and methanol for 2 days

each, the phthalocyanine products were extracted into

100 mL chloroform until the Soxhlet extracts were water

white. After evaporation of the chloroform, a green solid

(299 mg) was obtained. This green solid was ﬁrst puri-

ﬁed by gel-permeation chromatography (BioBeads SX1)

using toluene as eluent.

1

absorption peaks. 2. H NMR (CDCl₃): δ, ppm 9.74

(2H, s), 8.72 (4H, s), 8.69 (4H, s), 8.68 (4H, s), 4.55-

4.37 (24H, m), 2.12 (24H, m), 1.70 (24H, m), 1.53

(24H, m), 1.03 (36H, m). MS (MALDI): m/z calcd.

2093 (100%) [M + H]⁺, found 2093 (100%) [M + H]⁺

(pyridine coordination was often found in the MALDI

spectrum due to the use of pyridine in the eluent).

UV-vis (in pyridine): λ_max, nm 339 (Soret band, ε =

4.98 × 10⁴), 633, 699 (Q-band, ε = 3.68 × 10⁴).

The second green fraction coming out of the col-

umn was 1 (2,3,9,10,16,17,23,24-octakis(pentyloxy)iron-

phthalocyanine, 90 mg, 29%). Its purity (99%) was mea-

sured by HPLC with 2% pyridine in chloroform as the

eluent. Its UV spectrum was measured in neat pyridine

solution to get characteristic isolated phthalocyanine

absorption peaks. 1. ¹H NMR (CDCl₃): δ, ppm 8.66 (8H,

s), 4.42 (16H, t), 2.17 (16H, m), 1.50 (32H, m), 0.92 (24H,

t). MS (MALDI): m/z calcd. 1257.99 [M + H]⁺, found

1254.43 [M + H]⁺, 1274 [M + NH₃]⁺, 1291 [M + 2NH₃]⁺,

1215 [M + H–C₃H₇]⁺, 1185 [M + H–C₅H₁₁]⁺, 1170

[M + H–OC₅H₁₁]⁺. UV-vis (in pyridine): λ_max, nm 342

(Soret band, ε = 1.66 × 10⁴), 430, 598, 630, 660 (Q-band,

ε = 1.32 × 10⁴).

concluSIon

A new side-by-side iron phthalocyanine dimer (2)

was synthesized with peripheral substituents that render

it soluble in organic solvents. After puriﬁcation by gel

permeation chromatography and HPLC, a 12% yield

of the desired product in 98% pure form was obtained,

along with a major side product, the monomeric iron

phthalocyanine 1. There are a few novel aspects of this

new side-by-side pc dimer. First, this is a new compound

with full pentyloxy substitution (-O-C₅H₁₁), which gives

high solubility in organic solvents and a well behaved

NMR. Second, it was puriﬁed mainly by HPLC, which

is rarely used in pc puriﬁcation but has great potential.

The usage of the photodiode array detector with HPLC

data greatly helped the puriﬁcation process. Third, the

synthesis used NH₃as a protection gas to improve the

yield of the side-by-side pc dimer. With the addition of

pyridine, their NMR spectra of both monomer and dimer

became much clearer, and their UV-vis spectra indicated

a multiphase aggregation in DCM solution that was bro-

ken up as pyridine coordinated to the axial positions

The ﬁrst fraction was collected from the GPC col-

umn and the toluene was removed by rotary evapo-

ration (80 mg crude). It contained about 50% 2 and

50% 1. HPLC was used to do further puriﬁcation using

2% pyridine in chloroform as the eluent. A Phenom-

enex^TMLUNA normal phase silica analytical column

(250 × 4.6 mm, 100 Å) was used to identify and quan-

tify compounds 1 and 2. The ﬂow rate was set to 0.5

mL/second. Two major components were observed

at retention times of 5.7 and 6.7 min, and by UV-vis

measurements and mass spectrometry, these fractions

were identiﬁed as compounds 1 and 2, respectively.

Multiple injections were performed on the semi-prep

HPLC column. The ﬂow rate was set to 1 mL/second.

J. Porphyrins Phthalocyanines 2011; 15: 290–292

THe SynTHeSIS anD CHaraCTerIzaTIOn Of a SIDe-by-SIDe IrOn PHTHaLOCyanIne DIMer

291

of the iron pc. Oxidation of the Fe was associated with

13. Kleinwachter J and Hanack M. J. Am. Chem. Soc.

1997; 119: 10684–10695.

14. Asano Y and Kobayashi N. Tetrahedron Lett. 2004;

45: 9577–9580.

15. Dominguez DD, SnowAW, Shirk JS and Pong RGS.

J. Porphyrins Phthalocyanines 2001; 5: 582–592.

16. Lelievre D, Bosio L, Simon J, Andre JJ and Bense-

baa F. J. Am. Chem. Soc. 1992; 114: 4475–4479.

17. Li ZY. Ph.D. Thesis, University of Notre Dame:

South Bend, IN, 2001.

loss of coordinated pyridine. Considering that we may

need to use an axial ligand like pyridine to help graft

the cationic mixed-valence 2⁺(product of reaction I) to

a surface, this result indicates that the anionic mixed-

valence 2^-(product of reaction II) may be a better can-

didate as a mixed-valence compound for the molecular

QCA project. It remains to be seen whether the Fe-Pc-

pyridine linkage in molecules of 2⁺might be stable for

dry surface-bound molecules.

18. Li ZY and Lieberman M. Inorg. Chem. 2001; 40:

932–939.

Acknowledgements

19. Wang YL. and Lieberman M. IEEE Trans. Nano-

technol. 2004; 3: 368–376.

20. Janczak J and Kubiak R. Inorg. Chim. Acta 2003;

342: 64–76.

We thank Dr. David Alonso (Andrews University)

and his students Marco Allard, James Fox and Wil-

liam Hardesty for useful discussions. The National

Science Foundation supported this work ﬁnancially

(NSF CCF0403760).

21. Kennedy BJ, Murray KS, Zwack PR, Homborg H

and Kalz W. Inorg. Chem. 1985; 24: 3302–3305.

22. Kobayashi N, Koshiyama M, Funayama K, Osa T,

Shirai H and Hanabusa K. J. Chem. Soc.: Chem.

Commun. 1983: 913–914.

23. Kobayashi N, Kondo R, Nakajima S and Osa T.

J. Am. Chem. Soc. 1990; 112: 9640–9641.

24. Kobayashi N, Ishizaki T, Ishii K and Konami H.

J. Am. Chem. Soc. 1999; 121: 9096–9110.

25. Weitemeyer A, Kliesch H and Wohrle D. J. Org.

Chem. 1995; 60: 4900–4904.

Supporting information

Full separation and characterization information for 1

and 2 are given in the supplementary material. This mate-

rial is available free of charge via the Internet at http://

www.worldscinet.com/jpp/jpp.shtml.

rEFErEncES

1. Lent CS, Isaksen B and Lieberman M. J. Am. Chem.

Soc. 2003; 125: 1056–1063.

26. Kobayashi N, Muranaka A and Nemykin VN. Tetra-

hedron Lett. 2001; 42: 913–915.

2. Lent CS, Tougaw PD and Porod W. Proc. Workshop

on Physics and Computing, IEEE Computer Soci-

ety Press, 1994; 5–13.

3. Lieberman M, Chellamma S, Varughese B, WangY,

Lent C, Bernstein G, Snider G and Peiris F. Mol.

Electron. II, Ann. New York Acad. Sci. 2002; 960:

225–239.

4. Ellenbogen JC and Love JC. Proc. IEEE 2000; 88:

386–426.

5. Li M, Schnablegger H and Mann S. Nature 1999;

402: 393–395.

27. Kobayashi N, Miwa H and Nemykin VN. J. Am.

Chem. Soc. 2002; 124: 8007–8020.

28. Kimura M, Kuroda T, Ohta K, Hanabusa K, Shirai H

and Kobayashi N. Langmuir 2003; 19: 4825–4830.

29. Maya EM, Vazquez P and Torres T. Chem. Eur. J.

1999; 5: 2004–2013.

30. Nakagawa M, Rikukawa M, Sanui K and Ogata N.

Synth. Met. 1997; 84: 391–392.

31. Alonso C, Pascual MJ, Salomon AB, Abruna HD,

Gutierrez A, Lopez MF, Garcia-Alonso MC and

Escudero ML. J. Electroanal. Chem. 1997; 435:

241–254.

6. Loweth CJ, Caldwell WB, Peng XG, Alivisatos AP

and Schultz PG. Angew. Chem., Int. Ed. 1999; 38:

1808–1812.

32. Kobayashi N, Higashi Y and Osa T. Chem. Lett.

1994: 1813–1816.

7. Norris DJ and Vlasov YA. Adv. Mater. 2001; 13:

371–376.

8. Hu WC, Sarveswaran K, Lieberman M and

Bernstein GH. J. Vac. Sci. Technol., B 2004; 22:

1711–1716.

33. Kobayashi N, Lam H, Nevin WA, Janda P, Leznoff

CC, Koyama T, Monden A and Shirai H. J. Am.

Chem. Soc. 1994; 116: 879–890.

34. Yang J and Vandemark MR. Tetrahedron Lett. 1993;

34: 5223–5226.

9. Lowery MK, Starshak AJ, Esposito JN, Krueger PC

and Kenney ME. Inorg. Chem. 1965; 4: 128.

10. Pawlowski G and Hanack M. Synth. Stuttgart 1980:

287–289.

11. Caminiti R, Donzello MP, Ercolani C and Sadun C.

Inorg. Chem. 1999; 38: 3027–3029.

35. Ferencz A, Neher D, Schulze M, Wegner G, Viaene

L and De Schryver FC. Chem. Phys. Lett. 1995;

245: 23–29.

36. Hanack M, Gul A, Hirsch A, Mandal BK, Subrama-

nian LR and Witke E. Mol. Cryst. Liq. Cryst. 1990;

187: 365–382.

12. Ceyhan T, Altindal A, Ozkaya AR, Erbil MK,

Salih B and Bekaroglu Z. Chem. Commun. 2006:

320–322.

37. Kohn M. J. Am. Chem. Soc. 1951; 73: 480–480.

38. Freas RB and Campana JE. Inorg. Chem. 1984; 23:

4654–4658.

J. Porphyrins Phthalocyanines 2011; 15: 291–292

292

W. He anD M. LIeberMan

39. Bo SH, Tang DH, Liu XH and Zhen Z. Dyes Pigm.

2008; 76: 35–40.

40. Fukuda T, Homma S and Kobayashi N. Chem. Eur.

J. 2005; 11: 5205–5216.

41. Giribabu L, Kumar CV, Reddy VG, Reddy PY,

Rao CS, Jang SR, Yum JH, Nazeeruddin MK and

Gratzel M. Sol. Energy Mater. Sol. Cells 2007; 91:

1611–1617.

42. Vagin SI, Frickenschmidt A, Kammerer B and

Hanack M. Chem. Eur. J. 2005; 11: 6568–6573.

43. Bornsen KO and Mohr MD. Anal. Methods Instrum.

1995; 2: 158–160.

44. Knochenmuss R, Stortelder A, Breuker K and

Zenobi R. J. Mass Spectrom. 2000; 35: 1237–1245.

45. Wilkinson WR, Gusev AI, Proctor A, Houalla M

and Hercules DM. Fresenius J. Anal. Chem. 1997;

357: 241–248.

46. Chen MJ, Nunez L, Rathke JW and Rogers RD.

Organomet. 1996; 15: 2338–2344.

47. TseYH,SeymourP,KobayashiN,LamH,LeznoffCC

and Lever ABP. Inorg. Chem. 1991; 30: 4453–4459.

48. Edwards L and Gouterma M. J. Mol. Spectrosc.

1970; 33: 292.

63. Leznoff CC, Lam H, NevinWA, Kobayashi N, Janda

P and Lever ABP. Angew. Chem., Int. Ed. 1987; 26:

1021–1023.

64. Makarov S, Litwinski C, Ermilov EA, Suvorova

O, Roder B and Wohrle D. Chem. Eur. J. 2006; 12:

1468–1474.

65. De La Torre G, Martinez-Diaz MV and Torres T.

J. Porphyrins Phthalocyanines 1999; 3: 560–568.

66. Kobayashi N and Ogata H. Eur. J. Inorg. Chem.

2004: 906–914.

67. Kennedy BJ, Murray KS, Zwack PR, Homborg H

and Kalz W. Inorg. Chem. 1986; 25: 2539–2545.

68. Kuppers H, Eulert HH, Hesse KF, Kalz W and

Homborg H. Z. Naturforsch., B: J. Chem. Sci. 1986;

41: 44–47.

69. Kuppers H, Kalz W and Homborg H. Acta Crys-

tallogr., Sect. C: Cryst. Struct. Commun. 1985; 41:

1420–1423.

70. Lam H, Marcuccio SM, Svirskaya PI, Greenberg S,

Lever ABP, Leznoff CC and Cerny RL. Can. J.

Chem.: Rev. Can. Chim.1989; 67: 1087–1097.

71. Piechocki C and Simon J. J. Chem. Soc.: Chem.

Commun. 1985: 259–260.

49. Henrikss A and Sundbom M. Theor. Chim. Acta

1972; 27: 213.

72. Gok Y and Yildiz SZ. Polyhedron 1997; 16:

2335–2339.

50. Jerwin K and Wasgestian F. Spectrochim. Acta,

Part A.: Mol. and Biomol. Spectrosc. 1984; 40:

159–163.

51. Dodsworth ES, Lever ABP, Seymour P and Leznoff

CC. J. Phys. Chem. 1985; 89: 5698–5705.

52. Kane AR, Sullivan JF, Kenny DH and Kenney ME.

Inorg. Chem. 1970; 9: 1445.

73. Yeung YO, Liu RCW, Law WF, Lau PL, Jiang JZ

and Ng DKP. Tetrahedron 1997; 53: 9087–9096.

74. Gottlieb HE, Kotlyar V and Nudelman A. J. Org.

Chem. 1997; 62: 7512–7515.

75. Ishii K, Kobayashi N, Higashi K, Osa T, Lelievre D,

Simon J and Yamauchi S. Chem. Commun. 1999:

969–970.

53. Snow AW and Jarvis NL. J. Am. Chem. Soc. 1984;

106: 4706–4711.

76. Lever ABP. J. Porphyrins Phthalocyanines 1999; 3:

488–499.

54. Kobayashi N and Lever ABP. J. Am. Chem. Soc.

1987; 109: 7433–7441.

55. Nozawa T, Kobayashi N, Hatano M, Ueda M and

Sogami M. Biochim. Biophys. Acta 1980; 626:

282–290.

77. Lever ABP, Milaeva ER and Speier G. Phthalocya-

nines, Properties and Applications, Vol. 3, VCH:

NY, 1983; pp 1–70.

78. Wolberg A and Manassen J. J. Am. Chem. Soc.

1970; 92: 2982–2991.

56. Nyokong T, Gasyna Z and Stillman MJ. Inorg.

Chem. 1987; 26: 1087–1095.

57. Vancott TC, Rose JL, Misener GC, Williamson BE,

Schrimpf AE, Boyle ME and Schatz PN. J. Phys.

Chem. 1989; 93: 2999–3011.

79. Campbell RH, Heath GA, Hefter GT and McQueen

RCS. J. Chem. Soc.: Chem. Commun. 1983:

1123–1125.

80. Lever ABP and Wilshire JP. Inorg. Chem. 1978; 17:

1145–1151.

58. Janczak J and Kubiak R. Crystengcomm 2010; 12:

3599–3606.

81. Song L and Trogler WC. Angew. Chem. Int. Ed.

1992; 31: 770–772.

59. Stillman MJ and Thomson AJ. J. Chem. Soc.: Fara-

day Trans. II 1974; 70: 790–804.

60. Kobayashi H andYanagawaY. Bull. Chem. Soc. Jpn

1972; 45: 450.

61. Hush NS and Woolsey IS. Mol. Phys. 1971; 21: 465.

62. Kobayashi N, Fukuda T and Lelievre D. Inorg.

Chem. 2000; 39: 3632–3637.

82. van Nostrum CF, Picken SJ, Schouten AJ and

Nolte RJM. J. Am. Chem. Soc. 1995; 117: 9957–

9965.

83. Piechocki C. Ph.D. Thesis, University of Paris,

Paris, 1985.

84. BossardGE,AbramsMJ,DarkesMC,VollanoJFand

Brooks RC. Inorg. Chem. 1995; 34: 1524–1527.

J. Porphyrins Phthalocyanines 2011; 15: 292–292

Article Doi

The synthesis and characterization of a side-by-side iron phthalocyanine dimer

DOI: 10.1142/S1088424611003264

Source and publish data:

Authors:

Article abstract of DOI:10.1142/S1088424611003264

Full text of DOI:10.1142/S1088424611003264

Products guided by the article

R&D Labs maybe for 130996-04-2

Relevant to this article

Hot Product