Aspects of investigating scrambling in the synthesis of porphyrins: Different analytical methods

Article abstract of DOI:10.1016/j.tetlet.2005.06.123

Herein, we discuss the analyses and quantification of the different components in porphyrin mixtures, prepared from p-anisaldehyde, p-tolualdehyde, and 5-(4-bromophenyl)-dipyrromethane with acid catalysis, using NMR and HPLC. The advantages and disadvantages of these analytical methods are emphasized. Due to the similar size of a bromine atom and a methyl group it was possible to grow crystals suitable for X-ray crystallographic studies from a mixture of porphyrins, where the 4-position of the meso-phenyl rings was either substituted with methyl groups or bromine atoms. We also show that X-ray studies are inferior to NMR analysis for determining the components in a porphyrin mixture.

Full text of DOI:10.1016/j.tetlet.2005.06.123

Tetrahedron
Letters
Tetrahedron Letters 46 (2005) 5935–5939
Aspects of investigating scrambling in the synthesis of
porphyrins: different analytical methods
Christian B. Nielsen* and Frederik C. Krebs
Polymer Department, Risø National Laboratory, Frederiksborgvej 399, DK-4000 Roskilde, Denmark
Received 13 April 2005; revised 16 June 2005; accepted 22 June 2005
Available online 11 July 2005
Abstract—Herein, we discuss the analyses and quantification of the different components in porphyrin mixtures, prepared from
p-anisaldehyde, p-tolualdehyde, and 5-(4-bromophenyl)-dipyrromethane with acid catalysis, using NMR and HPLC. The advant-
ages and disadvantages of these analytical methods are emphasized. Due to the similar size of a bromine atom and a methyl group it
was possible to grow crystals suitable for X-ray crystallographic studies from a mixture of porphyrins, where the 4-position of the
meso-phenyl rings was either substituted with methyl groups or bromine atoms. We also show that X-ray studies are inferior to
NMR analysis for determining the components in a porphyrin mixture.
Ó 2005 Elsevier Ltd. All rights reserved.
The preparation of tetra meso-substituted porphyrins is
an active research area as such porphyrins are used as
molecular building blocks in areas such as molecular
electronics.1,2 Several methods exist for the preparation
of porphyrins with up to four different meso-substitu-
tents but common to all of these methods is that a por-
phyrinogen is formed, which is then oxidized to the
desired porphyrin. Acidolysis of the porphyrinogen
can lead to scrambling resulting in the isolation of a
mixture of up to six different porphyrins.3 Successful
attempts have been reported in the literature to suppress
the scrambling by using electron deficient, orthoesters,4
sterically hindered dipyrromethanes,5 or dipyrro-
methane itself, in the synthesis of porphyrins.5,6 A method
has also been published, where carbinols of dipyrro-
methanes were condensed with themselves to give the
corresponding porphyrinogens, which were then oxi-
dized to the porphyrins.7–9 This procedure resulted in
low or undetectable scrambling of the porphyrins dis-
cussed. In perspective, the conditions that have been
developed for suppressing scrambling in porphyrin syn-
thesis seem to be highly dependent on the porphyrin in
question and reliable detection and quantification of
the different porphyrins present in a mixture is impor-
tant for further investigations of scrambling in
porphyrinogens.
The purpose of the present work is to illustrate how to
quantify the distribution of scrambled porphyrins and
also present an X-ray study of the mixture of porphyrins
obtained from the synthesis of a meso-bromo-phenyl
and -tolyl substituted porphyrin. We have focused on
1
the use of H NMR, MALDI-TOF, and HPLC tech-
niques for quantifying the outcome of porphyrin synthe-
ses. These methods are discussed by applying them to
the product distribution obtained from porphyrin syn-
theses using 5-(4-bromophenyl)-dipyrromethane, p-anis-
aldehyde, and p-tolualdehyde. We also demonstrate that
a pure non-scrambled porphyrin is obtained when the
electron deficient 5-(4-cyanophenyl)-dipyrromethane is
reacted with p-anisaldehyde in accordance with previous
results.
The compounds isolated from the porphyrin synthesis
(Scheme 1) were characterized by NMR spectroscopy,
which revealed that 1 was isolated without the presence
of scrambling products in accordance with expectations
as we had used an electron deficient dipyrromethane.
However, 2 and 3 crystallized along with their scram-
bling products, which are clearly seen from the NH-part
1
(21H and 23H protons) of the H NMR spectrum (see
Fig. 1). MALDI-TOF of the porphyrin mixture result-
ing from the synthesis of 3 is depicted in Figure 2 and
we identified molecular masses corresponding to
porphyrins with zero, one, two, and three bromine
Keywords: Analysis of porphyrin mixtures; Analytical techniques;
NMR; HPLC; X-ray crystallography.
*
Corresponding author. Tel.: +45 46774798; fax: +45 46774791;
e-mail: christian.b.nielsen@gmail.com
0040-4039/$ - see front matter Ó 2005 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetlet.2005.06.123

5936
C. B. Nielsen, F. C. Krebs / Tetrahedron Letters 46 (2005) 5935–5939
R2
1. R2
CHO, TFA, RT, 30 min.
N
NH
NH
HN
N
2. DDQ, reflux, 1 h.
CH2Cl2
R1
R1
R1
NH
R2
CN, OMe 16% (1)
R1, R2
=
Br, OMe
Br, Me
(2) and scrambling
(3) and scrambling
Scheme 1. Syntheses of the trans-A2B2 substituted porphyrins discussed in the present work.
Figure 1. 1H NMR spectra of the NH region of 2 with scrambling
products (left) and 3 with scrambling products (right).
Figure 3. Four Lorentzian functions were fitted to the NH region of
the 1H NMR spectrum recorded for 3 with scrambling products.
The NH region in the 1H NMR spectrum of the mixture
of porphyrins resulting from the synthesis of 2 and 3 can
be fitted with four Lorentzian functions as shown in Fig-
ure 3. From the MALDI-TOF experiments, we know
that the mixture of 3 contained tetra-tolylporphyrin
(TTP). In independent experiments, we recorded the
1H NMR spectra of TTP and a tetra bromo-substituted
porphyrin and found that the NH protons had chemical
shifts of À2.76 and À2.86 ppm, respectively. Bromine
substitution thus causes an upfield shift of the NH pro-
tons. Using these results, we determined the distribution
of the bromine-containing porphyrins in the scrambling
mixture of 3: TTP occurred in 14% and the porphyrins
containing one, two, and three bromine atoms occurred
in 48%, 33%, and 5%, respectively, in the mixture. A
similar analysis of a porphyrin mixture has been
1
described,10 where the NH-signal in the H NMR spec-
trum was used to determine whether multiple products
were obtained from a porphyrin synthesis (the porphy-
rin mixture investigated in Ref. 10 had iodine and
hydrogen as substituents on the meso-phenyl rings
instead of Me and Br as in 2). However, the analysis
carried out in Ref. 10 was only quantitative and the
Figure 2. MALDI-TOF of 3 with scrambling products.
atoms. MALDI-TOF of the mixture of 2 (not shown) re-
vealed porphyrins with one, two, and three bromine
atoms.

C. B. Nielsen, F. C. Krebs / Tetrahedron Letters 46 (2005) 5935–5939
5937
R
1. BF3·Et2O, RT, 2 h.
NH
N
N
2. p-chloranil, reflux, 2 h.
+
R
CHO
NH
R
R
CH2Cl2
HN
R =
Br (4)
12%
CN (5) 15%
R
Scheme 2. Syntheses of the symmetrical porphyrins discussed in the present work.
distribution of porphyrins was not determined. In order
to investigate the limitations of using the resonances
from the NH protons in the H NMR spectrum of a
correlation with a slope of one is observed, as expected.
This experiment also serves as an illustration of the con-
centration dependence of the chemical shift of the NH
resonances: the concentration of 4 was varied from
0.40 to 2.77 mM and no change in the chemical shift
was observed. Thus the most severe limitation in the
NMR method is that it is not obvious from the struc-
tures when two different porphyrins will have NH reso-
nances that are clearly distinguishable from each other.
However, if the resonances can be clearly identified, the
NMR method seems to be a powerful tool for determin-
ing the composition of a porphyrin mixture. Of course,
the downfield region of the 1H NMR spectra of 1, 2, and
3 can be used for looking at the scrambling products ob-
tained from the oxidation of the porphinogens due to
the simplicity of the spectra. In porphyrins with a more
complex substitution pattern this part of the spectrum
can become complicated, whereas in the upfield region
above 0 ppm only the NH cavity protons have reso-
nances. Thus if 2D NMR spectroscopy is to be avoided
and only a ÔsimpleÕ 1H NMR spectrum is used for assess-
ing isomer purity, this region of the spectrum appears to
contain enough information.
1
porphyrin mixture, we first looked at mixtures of TPP
(tetraphenylporphyrin) and TTP prepared simply by
mixing pure TTP and TPP mixtures. However, the
NH protons in these two porphyrins had almost identi-
cal chemical shift values so it was not possible to deduce
anything about the composition of these mixtures. Simi-
larly, the NH protons in 5 (Scheme 2) had a chemical
shift of À2.87 ppm being almost identical to the value
of the NH protons in 4 (À2.86). Apparently, the electron
accepting or donating properties of the meso-substituent
is not the only factor that determines the chemical shift
of the NH resonances as the cyano group is a consider-
ably stronger electron acceptor than bromine.
Thus, from the structure alone it is not possible to deter-
mine whether the resonances for the NH protons will be
different for two porphyrins. A (control) validation of
1
the NH H NMR method was also carried out by pre-
paring mixtures with known concentrations of TTP
and 4 followed by determining the composition of the
mixtures by NMR. The correlation between the compo-
sition determined using the NMR method and the
ÔactualÕ composition is shown in Figure 4. A linear
The MALDI-TOF spectrum of 2 (not shown) only
revealed three components, whereas NMR clearly shows
four components. To elucidate this discrepancy, we ana-
lyzed the mixture of 2 by HPLC using gradient elution
with MeOH and THF. We monitored the absorbance
at 420 nm as a function of the retention time and
obtained the chromatogram depicted in Figure 5. By
integrating the peaks in the chromatogram we deter-
mined the distributions of the porphyrins to be 9%,
46%, 39%, and 6% of the porphyrins with 0, 1, 2, and
3 bromine atoms, respectively, which is in good agree-
ment with the product distribution from the NMR
results. In reproducing the HPLC result, we noted that
care should be taken with the solubility of the different
porphyrins in the mixture, as we obtained different
product distributions when some of the porphyrins
had precipitated from the HPLC sample. As four com-
ponents are observed in the mixture of 2 in both the
NMR and HPLC results, we can conclude that the
ÔmissingÕ porphyrin in the MALDI-TOF spectrum is
due to insufficient ionization efficiency. We also tried
to carry out HPLC analysis on 3 but we were unable
to find a method that resulted in the detection of the dif-
ferent porphyrins present in the mixture of 3. This is
Figure 4. NH part of the 1H NMR spectrum of mixtures of TTP and 4
(inset) and the correlation between the relative amount of 4 determined
from NMR and the actual contents of 4 in the mixtures.

5938
C. B. Nielsen, F. C. Krebs / Tetrahedron Letters 46 (2005) 5935–5939
Figure 5. HPLC chromatogram of 2 with scrambling products.
ascribed to the lack of functional groups in 3 that can
interact with the column stationary phase. In 2 there is
a varying amount of methyl ether groups in the mole-
cule, which can interact with the column stationary
phase through polar interactions. No polar groups are
present on the meso-substituent in 3. This finding sug-
gests a limitation of using HPLC for analyzing porphy-
rin mixtures. A further limitation of the HPLC method
(using absorbance as the detection method) is that differ-
ent porphyrins may have different extinction coefficients
meaning that the distribution found from a HPLC study
not only reflects the relative amount of the different por-
phyrins in the mixture but also differences in the extinc-
tion coefficients.
Figure 6. Structure and crystal packing (stereoview) of 3.
lengths between the aromatic core and the substituents
improved to reasonable values with C–Br bond dis-
˚
tances of 1.8666(23) and 1.8624(24) A, respectively,
and C–C bond distances of 1.8624(24) and
1.5957(124), respectively. The C–C bond distances are
perhaps a little larger than expected. This is ascribed
to the difficulty of modeling a disordered carbon atom
close to a heavy atom such as bromine. This is also
reflected in the U-values, which are large for the methyl
groups.
We were able to grow crystals suitable for X-ray crystal-
lography of 3 but not of 2, which is due to the similar
van der Waals radii of the bromine atoms and the
methyl groups.
It is only possible to obtain crystals from a porphyrin
mixture when the varying functional groups in the
molecular skeleton are of similar size. Single crystal X-
ray crystallography is not, in general, a viable method
for investigating scrambling in porphyrin synthesis.
Crystal structures can be obtained from powder X-ray
data, but this method is considerably more time con-
suming and is non-trivial. The purpose of including an
X-ray study of 3 is to have complementary information
The different analysis techniques give rise to differences
in the estimated bromine and methyl content for 3 as
shown in Table 1.
From this it is clear that while both NMR and HPLC
gives similar results, as seen from the data for 2, the
X-ray method, however, overestimates the bromine con-
tent in 3. The X-ray method is most likely limited due to
at least two different factors. Firstly it is difficult to cor-
1
to the H NMR data as it was not possible to analyze
the mixture of 3 using HPLC.
The X-ray studies on 3 revealed that the bromine/methyl
positions in the asymmetric unit were clearly disordered
exhibiting an electron density that was much lower than
that of a bromine atom and much higher than that of a
methyl group (Fig. 6). This was modeled as two overly-
ing mutually exclusive substituents, one bromine and
one methyl group that were refined freely with respect
to the atomic position and having anisotropic U-values.
The hydrogen atoms of the methyl group were fixed in a
standard methyl geometry and were riding on the
carbon atom. The sof was refined and this gave a slight
over representation of the bromine atoms. The bond
Table 1. A list of the bromine and methyl content obtained for 2 and 3
from each method of analysis
Technique
Br
Me/OMe
Compound 2
NMR
HPLC
1.59
1.58
2.41
2.42
Compound 3
NMR
X-ray
1.29
2.11
2.71
1.89

C. B. Nielsen, F. C. Krebs / Tetrahedron Letters 46 (2005) 5935–5939
5939
rectly model the occupancy when a heavy atom is close
to a light atom. This is also reflected in the relatively
long C–C bond length obtained. Secondly the result of
the single crystal X-ray analysis is on one single crystal
selected by the experimenter. The possibility of the crys-
tals being more well-formed when the bromine methyl
ratio is close to one cannot be excluded. Since the experi-
menter has a tendency to pick well-formed crystals this
bias could be problematic. If a more accurate bromine
methyl ratio should be obtained through the use of scat-
tering techniques, neutron scattering might be consid-
ered as a more reliable method.
Data Centre (CCDC No. 276494). Copies of this infor-
mation can be obtained free of charge via, www.ccdc.
cam.ac.uk. Experimental procedures for the synthesis
and for the X-ray analysis are provided as supplemen-
tary information. Supplementary data associated with
this article can be found, in the online version, at
doi:10.1016/j.tetlet.2005.06.123.
References and notes
1. Ashkenasy, G.; Cahen, D.; Cohen, R.; Shanzer, A.; Vilan,
A. Acc. Chem. Res. 2002, 35, 121–128.
2. Yoon, D. H.; Lee, S. B.; Yoo, K. H.; Kim, J.; Lim, J. K.;
Aratani, N.; Tsuda, A.; Osuka, A.; Kim, D. J. Am. Chem.
Soc. 2003, 125, 11062–11064.
3. Geier, G. R.; Lindsey, J. S. Tetrahedron 2004, 60, 11435–
11444.
4. Fox, S.; Hudson, R.; Boyle, R. W. Tetrahedron Lett. 2003,
44, 1183–1185.
5. Littler, B. J.; Ciringh, Y.; Lindsey, J. S. J. Org. Chem.
1999, 64, 2864–2872.
6. Cammidge, A. N.; Ozturk, O. Tetrahedron Lett. 2001, 42,
355–358.
7. Rao, P. D.; Dhanalekshmi, S.; Littler, B. J.; Lindsey, J. S.
J. Org. Chem. 2000, 65, 7323–7344.
8. Rao, P. D.; Littler, B. J.; Geier, G. R.; Lindsey, J. S.
J. Org. Chem. 2000, 65, 1084–1092.
In conclusion, we have described how a to perform an
analysis of porphyrin mixtures and determined the dis-
tribution of porphyrins arising from porphyrin synthesis
using NMR and HPLC analysis. Also the limitations of
these methods have been addressed. These relatively
simple methods can be applied in a routine fashion to
investigate in more detail the mechanisms responsible
for scrambling in asymmetric porphyrin synthesis by
analyzing the obtained product distribution. We have
carried out X-ray crystallographic studies on a mixture
of porphyrins obtained from condensing 5-(4-bromo-
phenyl)-dipyrromethane with p-tolulaldehyde.
9. Geier, G. R.; Callinan, J. B.; Rao, P. D.; Lindsey, J. S.
J. Porphyrins Phthalocyanines 2001, 5, 810–823.
10. Lee, C. H.; Li, F. R.; Iwamoto, K.; Dadok, J.; Bothnerby,
A. A.; Lindsey, J. S. Tetrahedron 1995, 51, 11645–
11672.
Supplementary data
Crystallographic data for the structural analysis of 3 has
been deposited with the Cambridge Crystallographic