Preparation and characterisation of hexamidine salts

Article abstract of DOI:10.1016/j.ijpharm.2015.07.071

Hexamidine diisethionate (HEX D) has been used in the personal care industry and in a number of over-the-counter (OTC) drug products as an antimicrobial agent since the 1950's. Recently, the compound has also been investigated for its beneficial effects on skin health. Surprisingly, there is only limited information describing the physicochemical properties of this compound in the literature. The objective of this work was therefore to conduct a comprehensive programme of characterisation of HEX D as well as its dihydrochloride salt (HEX H). HEX H was prepared from HEX D by a simple acid addition reaction. Both salts were characterised using Nuclear Magnetic Resonance (NMR), Differential scanning calorimetry (DSC), and Thermogravimetric analysis (TGA). A new high performance liquid chromatographic method was developed and validated for both compounds. The pH in aqueous solution as well as respective distribution coefficients between octanol and pH 7.4 buffer were also determined. Finally, solubility and short term stability studies were conducted in a range of solvents. NMR analysis confirmed the preparation of HEX H from HEX D. Thermal analysis indicated the melting points of HEX D and HEX H were 225 °C and 266 °C respectively. HPLC analysis confirmed the purity of both salts. Log D values at pH 7.4 were -0.74 for HEX D and -0.70 for HEX H respectively. The physicochemical properties of two HEX salts have been established using a range of analytical approaches. Detailed solubility and stability data have also been collated. This information will be useful in the design of novel formulations for targeted delivery of these compounds to the skin.

Full text of DOI:10.1016/j.ijpharm.2015.07.071

Accepted Manuscript
Title: Preparation and characterisation of hexamidine salts
Author: Nicola Parisi Paul J. Matts Rebecca Lever Jonathan
Hadgraft Majella E. Lane
PII:
S0378-5173(15)30090-9
DOI:
Reference:
http://dx.doi.org/doi:10.1016/j.ijpharm.2015.07.071
IJP 15075
To appear in:
International Journal of Pharmaceutics
Received date:
Revised date:
Accepted date:
3-6-2015
26-7-2015
28-7-2015
Please cite this article as: Parisi, Nicola, Matts, Paul J., Lever,
Rebecca, Hadgraft, Jonathan, Lane, Majella E., Preparation and
characterisation of hexamidine salts.International Journal of Pharmaceutics
http://dx.doi.org/10.1016/j.ijpharm.2015.07.071
This is a PDF file of an unedited manuscript that has been accepted for publication.
As a service to our customers we are providing this early version of the manuscript.
The manuscript will undergo copyediting, typesetting, and review of the resulting proof
before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that
apply to the journal pertain.

Preparation and characterisation of hexamidine salts
Nicola Parisi1, Paul J. Matts2, Rebecca Lever1, Jonathan Hadgraft1, Majella E. Lane1
1UCL School of Pharmacy
29-39 Brunswick Square
London
WC1N 1AX
United Kingdom
2Procter & Gamble Technical Centres Ltd,
London Innovation Centre,
Whitehall Lane,
Egham, Surrey
TW20 9NW
United Kingdom
Graphical abstract
a) Hexamidine diisethionate b) Hexamidine dihydrochloride
Abstract
Hexamidine diisethionate (HEX D) has been used in the personal care industry and in a number of
over-the-counter (OTC) drug products as an antimicrobial agent since the 1950’s. Recently, the
compound has also been investigated for its beneficial effects on skin health. Surprisingly, there is only
limited information describing the physicochemical properties of this compound in the literature. The

objective of this work was therefore to conduct a comprehensive programme of characterisation of
HEX D as well as its dihydrochloride salt (HEX H). HEX H was prepared from HEX D by a simple acid
addition reaction. Both salts were characterised using Nuclear Magnetic Resonance (NMR),
Differential scanning calorimetry (DSC), and Thermogravimetric analysis (TGA). A new high
performance liquid chromatographic method was developed and validated for both compounds. The
pH in aqueous solution as well as respective distribution coefficients between octanol and pH 7.4
buffer were also determined. Finally, solubility and short term stability studies were conducted in a
range of solvents. NMR analysis confirmed the preparation of HEX H from HEX D. Thermal analysis
indicated the melting points of HEX D and HEX H were 225°C and 266°C respectively. HPLC analysis
confirmed the purity of both salts. Log D values at pH 7.4 were -0.74 for HEX D and -0.70 for HEX H
respectively. The physicochemical properties of two HEX salts have been established using a range of
analytical approaches. Detailed solubility and stability data have also been collated. This information
will be useful in the design of novel formulations for targeted delivery of these compounds to the skin.
Key words: Hexamidine, salts, characterisation, preparation, pre-formulation, delivery
1.
Introduction
Hexamidine (HEX) is an aromatic diamidine and a strong organic base. Although primarily used
as the diisethionate salt (HEX D), it was firstly synthesised as the dihydrochloride (HEX H) and patented
by Ewins et al. (1939) for May & Baker Limited (U.K.). The company was interested in the trypanocidal
activity of the diamidines and the dihydrate of HEX H was subsequently demonstrated to be the most
potent of the group (Ashley et al., 1942). Antiprotozoal activity was demonstrated more than 50 years
later when Brasseur et al. (1994) used HEX D to treat two subjects affected by Acanthamoeba keratitis.
HEX D has also shown efficacy against Pseudomonas aeruginosa, Proteus, Escherichia coli,
Staphylococcus aureus and Tsukamurella paurometabolum (van Ketel, 1975; Granel et al., 1996). A
more recent in-vitro study demonstrated HEX D efficacy against a series of multi-drug resistant gram-

positive bacteria (Grare et al., 2010). Geratz et al. (1973) demonstrated the efficacy of HEX H dihydrate
as an enzyme inhibitor with Ki values of 1.9, 4.5 and 7.4 μM, trypsin, pancreatic kallikrein and thrombin
respectively. Enyedy et al. (2001) confirmed HEX inhibitory activity against thrombin (Ki value 224 nM)
and matriptase (Ki = 924 nM), but did not specify if the active was used as the free base or salt. Finally,
an in-vivo study investigated the effect of two HEX salts on nitric oxide synthase (NOS). Surprisingly,
while HEX D significantly decreased NOS activity, the tetrachloroplatinate (II) salt had no effect on NO
generation (Morgant et al., 1998).
A number of publications have focussed on the role of HEX as an anti-aging and moisturising
active in cosmetics and specifically the influence of HEX on various biomarkers of corneocyte maturity
and skin turnover. Kimball et al. (2012) speculated that HEX might attenuate the skin ageing process
because of its inhibitory activity on serine proteases associated with skin inflammation. Both skin
inflammation and abnormal lipid biosynthesis have been linked to skin ageing (McGrath et al., 2012).
Osborne et al. (2009) and Jarrold et al. (2010a) showed that when human skin equivalent cultures
were exposed to HEX, cholesterol, fatty acid and sphingolipid biosynthesis as well as cholesterol and
fatty acid uptake were downregulated while cholesterol efflux was upregulated. Jarrold et al. (2010b)
demonstrated that the application of a cosmetic moisturiser containing HEX, niacinamide and
palmitoyl-lysine-threonine significantly increased the number and size of mature corneocytes of the
facial stratum corneum of twenty female subjects. Significant thickening of the stratum corneum (SC)
as well as a reduction in transepidermal water loss of the volar forearm was reported for 36 female
subjects following treatment with a cream containing HEX and niacinamide (Kaczvinsky et al., 2010).
However these in vivo studies did not specify if the active was used as the free base or salt.
The safety of HEX and HEX D has been assessed by the Cosmetic Ingredient Review Expert
Panel (2007). The panel concluded that both actives are safe when used in cosmetics at concentrations
less than or equal to 0.10%. This opinion was subsequently confirmed by the European Parliament and
the Council of the European Union (2009) which fixed the maximum allowed concentration of HEX
and its salts in cosmetic products to 0.10%. However, several cases of allergic contact dermatitis have

been reported since HEX has been in use (Gougerot et al.1950; Sidi et al., 1969; van Ketel, 1975;
Robin,1978; Dooms-Goossens et al. 1989; Brand and Ballmer-Weber, 1995; Mullins, 2006;).
To date, HEX D has been used as a preservative in ~40 cosmetic products and in a number of
over-the-counter formulations (Cosmetic Ingredient Review Expert Panel, 2007). Surprisingly, there is
only a limited amount of information describing the physicochemical properties of HEX in the
literature (British Pharmacopoeia, 2015). The use of HEX H as an alternative salt to HEX D has also not
been explored. The objective, therefore, of the present work, was to undertake a comprehensive
programme of characterisation of HEX D and HEX H. In the longer term this information should assist
in the design of formulations which target this active more effectively to the skin.
2.
2.1 Materials
HEX D (Laboratoires Sérobiologiques, France) was a gift from Procter & Gamble (U.S.A.), while
Materials and Methods
HEX H was synthesized and purified in-house. Propylene glycol, polyethylene glycol 200, HPLC grade
isopropyl alcohol, trifluoroacetic acid (HPLC grade) and absolute ethanol were supplied by Fisher
Scientific (U.K.). HPLC grade solvents (acetonitrile, methanol, water), glycerol, isopropyl myristate, 1-
octanol, 2-ethylhexyl salicylate, 1 M hydrochloric acid solution and dimethyl sulfoxide-d6 were
provided by Sigma-Aldrich (U.K.). Dimethyl sulfoxide was supplied by VWR International (U.K.).
Propylene glycol monolaurate, Labrafac™ PG and Transcutol® P were received as gifts from Gattefossé
(France). 1,2-pentanediol was provided by Surfachem Group (U.K.). Dimethyl isosorbide (Arlasolve®)
was supplied by Croda International (U.K.). Oleic acid was provided by Fluka (U.K.). Miglyol® 812 N was
supplied by Sasol (Germany). Dipropylene glycol was provided by Acros Organics (Belgium). Phosphate
buffered saline was prepared using Dulbecco A tablets (Oxoid, U.K.).
2.2 Methods

Conversion of HEX D to HEX H
Approximately 50 mL of 1 M hydrochloric acid solution were heated at 100 ± 1°C using an
Ikamag® C-MAG HS 7 magnetic stirrer ceramic heating plate (IKA, Germany) equipped with an ETS-D5
electronic contact thermometer (IKA, Germany). HEX D was dissolved in the solution followed by
stirring of the mixture and cooling (15 min). The flask was subsequently placed on ice for 30 min to
allow recrystallisation of the product. Finally, crystals were recovered by means of vacuum filtration
and dried at room temperature. Hydrogen-1 and carbon-13 nuclear magnetic resonance (1H and 13C
NMR) spectroscopy were used to confirm the structure of the starting material and the product of the
reaction. All spectra were acquired in dimethyl sulfoxide-d6 on a Bruker Avance 400 MHz NMR
spectrometer (Bruker Corporation, U.S.A.) and processed using MestReNova® 9.0.1 (Mestrelab
Research, Spain).
Thermal analysis
The melting points of HEX D and HEX H were examined using thermogravimetric analysis (TGA)
and differential scanning calorimetry (DSC). TGA was performed using a Discovery TGA (TA
Instruments, U.S.A.) system. Each active was weighed in an open aluminium pan (TA Instruments,
U.S.A.) and then heated inside the Discovery TGA furnace. The starting temperature and the final
temperature were set to 25°C and 400°C, respectively, while the heating ramp was 10°C/min. A
nitrogen flow of 25 mL/min was supplied throughout the analysis in order to create an inert
atmosphere around the sample. A DSC Q2000 (TA Instruments, U.S.A.) system was used for the DSC
analysis. Each active was weighed in a hermetic aluminium pan (TA Instruments, U.S.A.) which was
subsequently sealed with a hermetic aluminium lid (TA Instruments, U.S.A.) using a Tzero™ press (TA
Instruments, U.S.A.). An empty hermetic aluminium pan (sealed with a hermetic aluminium lid) was
used as a reference. Both the sample and reference were heated from 40°C to 290°C, with a heating
ramp of10°C/min and a nitrogen flow of 50 mL/min.

UV, HPLC analysis and method validation
A Spectronic BioMate™ 3 UV/VIS spectrophotometer (Thermo Scientific, U.S.A.) was used to
carry out an UV scan of a solution of each active in HPLC grade water. The UV absorption spectrum
was acquired between 200 and 300 nm (step = 1 nm) in order to identify the wavelength at which the
absorption of light was specifically due to each active. The HPLC system consisted of a Hewlett-Packard
(U.S.A.) series 1100 quaternary pump, an Agilent Technologies (U.S.A.) series 1100 autosampler, a
Hewlett-Packard (U.S.A.) series 1100 system controller, an Agilent Technologies (U.S.A.) series 1100
degasser and an Agilent Technologies (U.S.A.) series 1100 UV detector. ChemStation® Rev.A.09.03
(Agilent Technologies, U.S.A.) software was used to analyse the data. HEX D was analysed with a Luna®
5 µm C8 150 × 4.60 mm reversed phase column (Phenomenex, U.K.) equipped with a universal HPLC
guard column (Phenomenex, U.K.) packed with a SecurityGuard™ C8 cartridge (Phenomenex, U.K.). The
mobile phase consisted of 75% v/v HPLC grade water (0.1% v/v HPLC grade trifluoroacetic acid) and
25% v/v HPLC grade acetonitrile. A Capcell Pak® MGIII 5 µm C18 250 × 4.60 mm reversed phase column
(Shiseido, Japan) was used to analyse HEX H. A universal HPLC guard column (Phenomenex, U.K.)
packed with a SecurityGuard™ C18 cartridge (Phenomenex, U.K.) was attached to the column. The
mobile phase consisted of 72% v/v HPLC grade water (0.1% v/v HPLC grade trifluoroacetic acid) and
28% v/v HPLC grade acetonitrile. For both HEX D and HEX H, the UV detector was set to 261 nm, the
flow rate to 0.7 mL/min and the column temperature to 35°C. The injection volume was set to 10 µL
for HEX D and 20 µL for HEX H. Linearity, specificity, accuracy, precision, lower limit of detection (LOD)
and lower limit of quantification (LOQ) of both methods were validated according to International
Conference on Harmonisation of Technical Requirements for Registration of Pharmaceuticals for
Human Use (2005).
pH and log Do/w determination
All pH measurements were taken using a SympHony® SB70P pH meter (VWR International,
U.K.) at 25 ± 1°C. Four solutions of each active in deionised water were tested (0.001, 0.01, 0.1 and 1

mM) with the pH of deionised water taken as the control. The method used to measure the log Do/w
of HEX D and HEX H was adapted from OECD guidelines (2006). 1-octanol was mutually saturated with
PBS (pH = 7.4 ± 0.5 at 25°C) by slow-stirring for 48 h at 25 ± 1°C. The system was allowed to equilibrate
in a separation funnel for 24 h. Two solutions of known concentrations of HEX D or HEX H in PBS
saturated with 1-octanol (pH = 7.4 ± 0.5 at 25°C) were prepared. Solutions were mixed in different
proportions (1:1, 2:1 and 1:2) with 1-octanol saturated with PBS (pH = 7.4 ± 0.5 at 25°C), placed in
glass test tubes sealed with Parafilm® and allowed to rotate on a rotor for 24 h at 25 ± 1°C. The two-
phase systems were then left to stand and equilibrate for 48 h at the experimental temperature. At
the end of the equilibration period, both phases were sampled with dilution where necessary.
Amounts of HEX D and HEX H were measured by HPLC and used to calculate the log Do/w (pH = 7.4) as
follows:
[
]
ꢂꢃꢄꢅꢆꢇ
ꢈꢉꢊꢋꢌꢈꢍ
log Dꢀ/ꢁ = log
(Equation
ꢎ
ꢗ
ꢂꢃꢄꢅꢆꢇ
ꢏꢋꢊꢐꢑ (ꢒꢓ ꢔ ꢕ.ꢖ)
1)
Solubility and stability studies
For solubility determination an excess amount of active was added to each solvent in a glass
test tube containing a Teflon®-coated magnetic stir bar. The test tube was sealed with Parafilm® and
placed in a SUB 28 thermostatically controlled water bath (Grant Instruments, U.K.) equipped with a
Telesystem HP 15 submersible magnetic stirrer (Variomag®-USA, U.S.A.). The system was allowed to
stir and equilibrate for 48 h at 32 ± 1°C to obtain a saturated solution. After the 48 h period, a sample
was withdrawn from the test tube and centrifuged at 13200 rpm for 15 min at 32 ± 1°C in an Eppendorf
5415R centrifuge (Eppendorf, Germany). Finally, the supernatant was suitably diluted and the
concentration of the active was determined by HPLC. Stability of HEX D and HEX H in several solvents
and binary solvent systems was investigated for 120 h at 32 ± 1°C. A solution of known concentration
of active was prepared and placed in a screw top glass test tube with a stir bar. The sample was sealed

and allowed to stir for 120 h at 32 ± 1°C as for solubility studies and aliquots were removed at 0, 24,
48, 72, 96 and 120 h. Following sample dilution the concentration of the active was determined by
HPLC.
3.
Conversion of HEX D to HEX H
The hypothetical double displacement reaction between HEX D and HCl is shown in Equation
Results and discussion
2.
(
)
(
)
ꢘ
CꢘꢙHꢘꢚNꢛOꢘ CꢘHꢚOꢛS ꢘ + 2HCl → CꢘꢙHꢘꢚNꢛOꢘ HCl ↓ + 2CꢘHꢚOꢛS
(2)
The 1H NMR spectrum of HEX D in dimethyl sulfoxide-d6 is shown in Figure 2a. The dimethyl
sulfoxide-d6 quintuplet at 2.50 ppm was used as a reference to scale the x-axis of the spectrum. A
water singlet at 3.30 ppm reflects the hygroscopicity of dimethyl sulfoxide-d6 which readily absorbs
moisture from the atmosphere and glassware (Gottlieb et al., 1997).
Figure 2a shows two triplets at 2.68 and 3.65 ppm which are assigned to the methylene
hydrogens of the isethionate anion (g and h respectively), while the singlet at 4.47 ppm is assigned to
the hydroxyl group (i). Interestingly, these peaks are not present in the 1H NMR spectrum of the HEX
H crystals, while those for the HEX moiety of the molecule are evident (Figure 2b). In Figure 2b, the
water signal is more intense than for Figure 2a and cannot be attributed solely to the moisture
absorbed by the dimethyl sulfoxide-d6. This strong signal may reflect residual aqueous reaction
medium or water of crystallisation which becomes trapped inside the crystals during the
recrystallisation process. Further confirmation of HEX H as the product is provided by comparison of
the 13C spectrum of HEX D with that of HEX H (Figures 3a and 3b ). The DMSO-d6 septuplet at 39.52
ppm was used as a reference to scale the x-axis of both spectra. Two singlets at 53.69 and 57.66 ppm
(Figure 3a) are assigned to the methylene carbons of the isethionate anion (9 and 10 respectively).
Again, these peaks are not present in the 13C NMR spectrum of the reaction product (Figure 3b).

Thermal analysis
The results of the TGA and the DSC analysis of HEX D are shown in Figure 4. TGA is a well-
established method for the characterisation of materials and is particularly useful in determining loss
of water molecules and compound degradation temperatures (Coats and Redfern, 1963). There is no
weight loss of HEX D between 25°C and 290°C (Figure 4a). However, degradation occurs between
300°C and 375°C and only ~ 12% of the initial weight of HEX D remains at 400°C. For DSC analysis, two
endothermic events were observed; the first has an onset temperature of 176.62°C and the second is
224.86°C. It may be hypothesised that these two peaks reflect the melting of two different crystal
structures of HEX D. Considering that pentamidine diisethionate, the lower homologue of HEX
diisethionate, exists in at least four crystalline forms (Steele, 1990; Chongprasert et al., 1998), the
possibility of multiple polymorphs of HEX D was expected. Fucke et al. (2008) identified ten anhydrous
and two dihydrate polymorphic forms of HEX D. Furthermore, the authors confirmed that HEX D does
not directly melt but undergoes a phase transition (Personal communication Fucke, 2015). This
suggests that the first endotherm in Figure 4a is the phase transition from a low-temperature form to
the stable high-temperature crystal form which melts at 224.86°C. The corresponding results for DSC
and TGA analyses of HEX H are shown in Figure 4b.
HEX H exhibits 11.1% weight loss between 25°C and 100°C and single-stage degradation
between 265°C and 350°C. Thus only ~7% of the initial weight of HEX H remains at 400°C. The initial
weight loss may be attributed to the evaporation of water from the sample. This is consistent with the
1H NMR spectrum of HEX H (Figure 3) and the presence of water of crystallisation. The water content
of 11.1% gives a stoichiometric ratio of three molecules of water per molecule of HEX H indicating the
salt was recrystallised in its trihydrate form. DSC analysis of HEX H shows three endothermic events
(Figure 4b). The first occurs between 40°C and 120°C, and represents the loss of water of crystallisation
already observed in the TGA curve. The second peak has on onset temperature of 223.2°C. This value
is very close to the melting point of the stable high-temperature crystal form of HEX D (Figure 4a). It
might be speculated that this second endotherm was the melting of residual HEX D which was not

converted to HEX H and remains as an impurity at the end of the conversion reaction. Finally, the third
sharp endothermic event with an onset temperature of 265.5°C is presumed to be the melting point
of HEX H.
UV and HPLC analysis and method validation
HEX D and HEX H exhibited a suitable UV peak for analysis at 261 nm. For the HPLC analysis
calibration curves (ranging from 0.5 µg/mL to 20 µg/mL) were constructed. The linearity for both
methods was confirmed by the correlation coefficient (r2) which was equal to 0.99 across the
experimental range. There were no interfering peaks at the retention times of the analytes which were
5. 1 min for HEX D and 8.2 min for HEX H. Recovery of each compound within the range from 90% to
110% was achieved. In addition, the %RSD for the intra-day and inter-day precision were below 5%
and 10% respectively, thus demonstrating the repeatability of the proposed methods. The LOD and
LOQ for HEX D were 0.54 µg/mL and 1.64 µg/mL. The values obtained for HEX H, were 0.40 µg/mL for
the LOD of and 1.21 µg/mL for the LOQ. These values are also lower than values previously reported
for HPLC analysis of HEX D (Taylor et al., 1983; De Bukanski and Masse, 1984).
pH in aqueous solution and log Do/w at pH = 7.4
Solutions of HEX D and HEX H in deionised water were as expected slightly acidic (pH ranging
from 6.3 to 6.4). The log Do/w at pH = 7.4 and 25 ± 1°C and the recovery of HEX D and HEX H are
reported in Table 1. Both actives showed a negative log Do/w, with HEX D having a significantly lower
value than HEX H (t-test, p < 0.01).
Solubility and stability
The solubility at 32 ± 1°C of HEX D and HEX H in a range of different solvents is shown in Figures
5a and 5b. Data for solvents in which both actives had solubility > 1 mg/mL are pooled in Figure 5a

while those in which they had solubility < 1 mg/mL are presented in Figure 5b. The only exception to
this is TC which is included in both figures. HEX H, in fact, had a solubility of 2.00 mg/mL in TC while
the value for HEX D was only 0.37 mg/mL.
Both actives exhibited highest solubility in DMSO compared with all the other solvents
studied; both actives were also soluble in PG, glycerol and methanol, sparingly soluble in 1,2-PENT and
PEG 200 and only slightly soluble in PBS, ethanol, DPG and TC (HEX H only). In addition, HEX D was
soluble in water, while HEX H was only sparingly soluble in water. The solubility of HEX D and HEX H
in water was fifteen and ten times, respectively, higher than that in PBS (pH = 7.4). Considering that
the pH of water was 6.36 and that the pKa of the amidino group of HEX is 11, the increase in pH resulted
in a lower ionisation and, as a result, in a lower solubility of the actives in PBS. This effect of pH on
solubility is commonly accepted and Avdeef (2007) has recently reviewed how it affects sparingly
soluble ionisable drugs. The presence of other ions and components of the buffer is also expected to
influence the solubility values obtained. For example, it is possible that phosphate anions may interact
with hexamidine cations, precipitate them and reduce hexamidine concentration in solution. As no
information is available in the literature on phosphate salts of HEX this is an area which deserves
further investigation.
With the exception of TC and IPA, both HEX D and HEX H were practically insoluble in all other
solvents studied. HEX H in particular, was so poorly soluble in 2-EHSAL, IPM and LABR that its solubility
was below the LOQ (1.21 µg/mL) for HPLC analysis. The percentage of HEX D recovered after 24, 48,
72, 96 and 120 h at 32 ± 1°C in a series of solvents and selected binary solvent systems is shown in
Table 2. The results summarised in Table 2 indicate that HEX D exhibits some instability in water. At
24 h recovery was 86.1 ± 3.6 % but there was no further degradation. Conversely, HEX D did not
undergo degradation in the other solvents and binary solvent systems tested. Less than 8 % loss was
observed after 120 h in PBS, PG, PEG 200, glycerol, PG:PGML (50:50) and DMSO:Methanol (50:50).The
results of the stability studies of HEX H in the same solvents and binary solvent systems seen for HEX
D are presented in Table 3. HEX H did not show any stability issues (Table 3) and less than 5 % loss was

observed after 120 h in water, PBS, PG, PEG 200, glycerol, PG:PGML (50:50) and DMSO:Methanol
(50:50).
4.
Conclusions
The selection of an active ingredient and the characterisation of its physicochemical
properties is arguably the most important stage in the preformulation design of a topical. All available
information about HEX and its salts was identified and reviewed. Although HEX D is the active that is
currently used in personal care and pharmaceutical formulations, its dihydrochoride salt, HEX H also
appears to be a suitable candidate molecule for delivery to the skin. We have confirmed that HEX H
has a lower MW than HEX D but a higher melting point. Thermal analysis also confirmed that HEX D
exists in different crystal forms and revealed that HEX H had recrystallised as a trihydrate during the
conversion process. The measurement of the pH of the solutions of HEX D and HEX H in deionised
water demonstrated that both salts are very weakly acidic. New HPLC analytical methods for the
quantification of HEX D and HEX H were developed and validated. The solubility of HEX D and HEX H
was studied in 19 solvents and both actives were found to be more soluble in those solvents having
polar properties. The stability of HEX D and HEX H in solution and in a limited number of combinations
of selected excipients was also evaluated. Overall, the findings are expected to be useful in the rational
design of new formulations for both actives.
Acknowledgements: NP is very grateful for financial support from Procter & Gamble, U.K.
References
Ashley, J.N., Barber, H.J., Ewins, A.J., Newbery, G. and Self, A.D.H., 1942. A chemotherapeutic
comparison of the trypanocidal action of some aromatic diamidines. J. Chem. Soc. pp.103-116.

Brand, C.U. and Ballmer-Weber, B.K., 1995. Contact sensitivity to 5 different ingredients of a topical
medicament (Imacort® cream). Contact Dermatitis 33 (2), pp.137.
Brasseur, G., Favennec, L., Perrine, D., Chenu, J.P. and Brasseur, P., 1994. Successful treatment of
Acanthamoeba keratitis by HEX. Cornea 13 (5), pp.459-462.
British Pharmacopoeia Commission, 2014. HEX isetionate. In: British Pharmacopoeia 2015 Volume I.
London: The Stationery Office, pp.1136-1138.
Chongprasert, S., Griesser, U.J., Bottorff, A.T., Williams, N.A., Byrn, S.R. and Nail, S.L., 1998. Effects of
freeze-dry processing conditions on the crystallization of pentamidine isethionate. J. Pharm. Sci. 87
(9), pp.1155-1160.
Coats, A.W. and Redfern, J.P., 1963. Thermogravimetric analysis. A review. Analyst, 88 (1053), pp.906-
924.
Cosmetic Ingredient Review Expert Panel, 2007. Final report on the safety assessment of HEX and HEX
diisethionate. Int. J. Toxicol. 26 (Suppl. 3), pp.79-88.
De Bukanski, B.W. and Masse, M.O., 1984. Analysis of HEX, dibromoHEX, dibromopropamidine and
chlorhexidine in cosmetic products by high-performance liquid chromatography. Int. J. Cosmet. Sci. 6
(6), pp.283-292.

Dooms-Goossens, A., Vandaele, M., Bedert, R. and Marien, K., 1989. HEX isethionate: a sensitizer in
topical pharmaceutical products and cosmetics. Contact Dermatitis 21 (4), pp.270.
Enyedy, I.J., Lee, S., Kuo, A.H., Dickson, R.B., Lin, C. and Wang, S., 2001. Structure-based approach for
the discovery of bis-benzamidines as novel inhibitors of matriptase. J. Med. Chem., 44 (9), pp.1349-
1355.
Ewins, A.J., Barber, H.J., Newbery, G., Ashley, J.N. and Self, A.D.H., May & Baker Limited, 1939. Process
for the preperation of amidine derivatives. United Kingdom. Patent 507,565 (A).
Fucke, K., 2015. RE: Poster - Polymorphic study of the model system HEX diisethionate. [email]
(Personal communication, 3 March 2015).
Fucke, K., Toebbens, D., Kahlenberg, V. and Griesser, U.J., 2008. Polymorphic study of the model
system HEX diisethionate. Acta Crystallogr. A 64 (Suppl. 1), pp.C447-C448.
Geratz, J.D., Whitmore, A.C., Cheng, M.C.F. and Piantadosi, C., 1973. Diamidino-,-
diphenoxyalkanes. Structure-activity relations for the inhibition of thrombin, pancreatic kallikrein, and
trypsin. J. Med. Chem. 16 (9), pp.970-975.
Gougerot, H., Tabernat, J., Raufast and Gascoin, 1950. Eczema généralisé par sensibilisation à
l'hexomedine. Bulletin de la Société Française de Dermatologie et Syphiligraphie, 57 (3), pp.271.

Granel, F., Lozniewski, A., Barbaud, A., Lion, C., Dailloux, M., Weber, M. and Schmutz, J., 1996.
Cutaneous infection caused by Tsukamurella paurometabolum. Clin. Infect. Dis. 23 (4), pp.839-840.
Grare, M., Dibama, H.M., Lafosse, S., Ribon, A., Mourer, M., Regnouf-de-Vains, J.B., Finance, C. and
Duval, R.E., 2010. Cationic compounds with activity against multidrug-resistant bacteria: interest of a
new compound compared with two older antiseptics, HEX and chlorhexidine. Clin. Microbiol. Infect.
16 (5), pp.432-438.
International Conference on Harmonisation of Technical Requirements for Registration of
Pharmaceuticals for Human Use, 2005. Validation of analytical procedures: text and methodology Q2
(R1).
ICH
harmonised
tripartite
guideline
[online].
Available
at:
<http://www.ich.org/fileadmin/Public_Web_Site/ICH_Products/Guidelines/Quality/Q2_R1/Step4/Q
2_R1__Guideline.pdf> [Accessed 13 January 2015].
Jarrold, B.B., Tiesman, J., Robinson, M., Binder, R. and Osborne, R., 2010a. HEX, a protease inhibitor,
promotes stratum corneum lipid biomarkers in vitro. J. Am. Acad. Dermatol. 62 (3, Suppl. 1), pp.AB2.
Jarrold, B.B., Kaczvinsky, J., Matts, P.J. and Osborne, R., 2010b. Use of a cosmetic moisturizer promotes
corneocyte maturity. J. Am. Acad. Dermatol. 62 (3, Suppl. 1), pp.AB62.
Kaczvinsky, J., Li, J., Crowther, J., Mirkovic, S. and Janson, W., 2010. Effect of topical antiaging products
on stratum corneum thickness and barrier integrity. J. Am. Acad. Dermatol. 62 (3, Suppl. 1), pp.AB25.

Kimball, A.B., Grant, R.A., Wang, F., Osborne, R. and Tiesman, J.P., 2012. Beyond the blot: cutting edge
tools for genomics, proteomics and metabolomics analyses and previous successes. Br. J. Dermatol.
166 (Suppl. S2), pp.1-8.
McGrath, J.A., Robinson, M.K. and Binder, R.L., 2012. Skin differences based on age and chronicity of
ultraviolet exposure: results from a gene expression profiling study. Br. J. Dermatol. 166 (Suppl. 2),
pp.9-15.
Morgant, G., Viossat, B., Roch-Arveiller, M., Prognon, P., Giroud, J.P., Lancelot, J.C., Robba, M. and
Huy, D.N., 1998. In vivo nitric oxide synthase inhibitors can be deprived of this activity: unexpected
influence of the tetrachloroplatinate(II) counteranion. Crystal structures of bis(S-methyl-
isothiouronium)-N,N'-bis(3-guanidinopropyl)piperazinium and hexamidinium tetrachloroplatinates(II)
salts. Metal-Based Drugs, 5 (3), pp.127-137.
Mullins, R.J., 2006. Systemic allergy to topical HEX. Med. J. Aust. 185 (3), pp.177.
Organisation for Economic Cooperation and Development, 2006. Test No. 123: partition coefficient (1-
octanol/water): slow-stirring method. OECD guidelines for the testing of chemicals, section 1: physical-
chemical properties. Paris: OECD Publishing, pp.1-15.
Osborne, R., Mullins, L.A. and Jarrold, B.B., 2009. Understanding metabolic pathways for skin anti-
aging. J. Drugs Dermatol. 8 (Suppl. 7), pp.s4-s7.
Robin, J., 1978. Contact dermatitis to HEX. Contact Dermatitis, 4 (6), pp.375-376.

Sidi, E., Bourgeois Spinasse, J. and Arouete, J., 1969. Quelques causes d'eczéma d'origine
médicamenteuse. Revue Française d'Allergie, 9 (3), pp.179-182.
Steele, G., 1990. An observation of polymorphism in pentamidine isethionate. J. Pharm. Pharmacol.
42 (2), pp.121-122.
Taylor, P., Braddock, P.D. and Ross., S. 1983. Quantitative determination of HEX isethionate in
pharmaceutical preparations by high-performance liquid chromatography. J. Pharm. Sci. 72 (12),
pp.1477-1478.
van Ketel, W.G. 1975. Allergic contact eczema by Hexomedine®. Contact Dermatitis, 1 (5), pp.332.

List of Figures
List of Figures
Figure 1. Chemical structures of (a) HEX D and (b) HEX H
Figure 2. 1H NMR spectrum of (a) HEX D and (b) HEX H in dimethyl sulfoxide-d6
Figure 3. 13C NMR spectrum of (a) HEX D and (b) HEX H in dimethyl sulfoxide-d6
Figure 4. TGA and DSC analysis of (a) HEX D and (b) HEX H
Figure 5. Solubility data for HEX D and HEX H at 32°C (a) Solubility > 1mg/mL (b) Solubility < 1mg/mL with exception of Transcutol® P which is shown in
both figures (n> 3; Mean±S.D.)
Figure 1. Chemical structures of (a) HEX D and (b) HEX H
Figure 1. Chemical structures of (a) HEX D and (b) HEX H
Figure 4. TGA and DSC analysis of (a) HEX D and (b) HEX H
Figure 4. TGA and DSC analysis of (a) HEX D and (b) HEX H
Figure 5. Solubility data for HEX D and HEX H at 32°C (a) Solubility > 1mg/mL (b) Solubility < 1mg/ml with exception of Transcutol™ which is shown in both
figures (n≥ 3; Mean±S.D.)

Figure 5. Solubility data for HEX D and HEX H at 32°C (a) Solubility > 1mg/mL (b) Solubility < 1mg/ml with exception of Transcutol™ which is shown in both
figures (n≥ 3; Mean±S.D.)

NH
a)
NH2
O
O
H2N
SO3H
NH
HO
2
NH
b)
NH2
O
O
H2N
HCl
NH
2
Figure 1

Figure 2

Figure 3

Figure 4

Figure 5

Table 1 Log Do/w at pH = 7.4 and 25 ± 1°C and recovery of HEX D and HEX H (n=9; mean ± SD)
Active
HEX D
HEX H
log Do/w
Recovery (%)
101.2 ± 2.7
101.5 ± 1.6
− 0.74 ± 0.02
− 0.70 ± 0.02

Table 2. Recovery (%) of HEX D in a series of solvents and binary solvent systems after 24, 48, 72, 96 and 120 h at 32 ± 1°C (3≤n≤4; mean ± SD)
PG:PGML
(50:50)
DMSO:Methanol
(50:50)
Time (h)
Water
100.0
PBS
PG
PEG 200
100.0
Glycerol
100.0
0
24
48
72
96
120
100.0
100.0
100.0
100.0
86.1 ± 3.6
80.7 ± 6.5
83.7 ± 1.5
82.5 ± 3.8
82.6 ± 4.4
99.27 ± 4.42
100.3 ± 4.5
99.9 ± 2.5
100.3 ± 2.2
102.1 ± 1.6
99.83 ± 1.96
103.3 ± 2.9
98.8 ± 3.4
99.2 ± 3.7
104.0 ± 7.9
94.36 ± 2.34
91.8 ± 7.8
87.8 ± 3.6
92.4 ± 1.7
93.0 ± 4.2
97.64 ± 2.83
98.4 ± 3.1
99.5 ± 2.0
98.3 ± 4.0
98.3 ± 3.4
94.78 ± 5.83
95.1 ± 4.2
95.4 ± 1.6
93.8 ± 1.6
94.6 ± 3.1
93.57 ± 2.34
99.3 ± 1.7
98.5 ± 0.8
98.6 ± 1.8
102.1 ± 3.0

Table 3. Recovery (%) of HEX H in a series of solvents and binary solvent systems after 24, 48, 72, 96 and 120 h at 32 ± 1°C (n=4; mean ± SD)
PG:PGML
(50:50)
DMSO:Methanol
(50:50)
Time (h)
Water
100.0
PBS
PG
PEG 200
100.0
Glycerol
100.0
0
24
48
72
96
120
100.0
100.0
100.0
100.0
98.4 ± 1.8
99.7 ± 2.9
99.8 ± 2.7
100.0 ± 2.5
100.3 ± 1.9
101.2 ± 1.3
98.9 ± 1.5
100.3 ± 2.1
100.7 ± 0.2
99.1 ± 2.4
99.4 ± 2.5
101.1 ± 2.7
100.8 ± 2.7
100.6 ± 3.2
101.2 ± 2.1
99.6 ± 0.6
98.4 ± 1.2
98.5± 1.9
96.6 ± 2.5
95.1 ± 1.5
97.0 ± 0.3
101.0 ± 1.1
99.8 ± 2.1
99.5 ± 2.4
101.4 ± 2.4
101.5 ± 3.7
101.7 ± 2.6
99.9 ± 2.9
101.1 ± 4.3
99.6 ± 2.4
93.1 ± 2.3
96.3 ± 2.6
98.3 ± 1.9
97.2 ± 1.6
102.1 ± 1.3