G Model

POC-3474; No. of Pages 5 ARTICLE IN PRESS

Progress in Organic Coatings xxx (2014) xxx–xxx

Contents lists available at ScienceDirect

Progress in Organic Coatings

journal homepage: www.elsevier.com/locate/porgcoat

Effect of hardener variation on protective properties

of polyurethane coating夽
Ewa A. Papaj, Douglas J. Mills ∗ , Sina S. Jamali
School of Science and Technology, The University of Northampton, St George’s Avenue, Northampton NN2 6JD, UK

a r t i c l e i n f o a b s t r a c t

Article history: Polyurethane is a two component coating commonly used in high performance anti-corrosion protective
Available online xxx systems. The two components are normally a polyacrylic resin and a polyisocyanate hardener. The main
purpose of this study was to investigate the effect of hardener variation on the protective properties of
Keywords: polyurethane coating. Apart from theoretical interest, there is a practical interest to this work because, for
Hardener ratio various reasons, in the field the correct ratio of base to hardener is not always followed. Another aim was to
Polyurethane coating
discover changes in mechanical properties such as abrasion and adhesion, and also find out differences in
Electrochemical properties
the chemical structure and crosslink density. Hence unpigmented polyurethane resin with hardener:base
Mechanical properties
Corrosion protection
ratios of 0.75:1, 1:1 and 1.5:1 were applied on mild steel panels. Detached coatings were obtained from
casting on polypropylene sheets. To evaluate anticorrosive properties attached coatings were immersed
in 3% sodium chloride for 1008 h and monitored using EIS and DC resistance tests. Abrasion resistance was
measured by the Taber test. Wet adhesion was studied by immersing in 3% NaCl for 5 and 7 days, and then
using the pull-off method. The detached coatings were used to evaluate chemical structure using FTIR
spectroscopy, and to look at the physico-mechanical properties using a DMTA test. Electrochemical results
showed that polyurethane coatings with a higher hardener ratio have a higher electrical (ionic) resistance.
This was attributed to a more tightly cross-linked structure. In addition, polyurethane coatings with a
higher hardener ratio have a better abrasion resistance. The lowest adhesion was given by the coating with
the stoichiometric hardener ratio. The work raised some interesting questions on the relation between
adhesion and corrosion resistance. The FTIR results indicated that polyurethane coatings with the lower
hardener ratio have fewer urethane groups than coatings with the higher hardener ratio. Furthermore
results confirmed that the Electrochemical Impedance Spectroscopy and DC resistance are good methods
to investigate how the protection ability of the coatings depends on both the nature of the coating and
on time.
© 2014 Elsevier B.V. All rights reserved.

1. Introduction common polyacrylics have hydroxyl groups as a part of their chem-

ical structure. Paint manufacturers always provide users with a
Organic coatings are the most popular method to protect mate- recommended base to hardener mixing ratio to produce optimized
rials against corrosion. Among such coatings polyurethane is one coating properties that is required for a given application. This ratio
of the most commonly used because it has excellent weathering is most often obtained empirically by the manufacturer or is sim-
and adhesion properties as well as chemical and abrasion resis- ply based on the equimolar ratio of hydroxyl (OH) and isocyanate
tance [1,2]. Most anti-corrosive polyurethane polymers consist of (NCO) groups [3]. There is no information easily available about
a polyacrylic resin (base) and a polyisocyanate hardener. The most properties of coatings when the hardener to resin ratio is different
to that recommended. In practice there are several cases where
a different hardener ratio in a coating may occur. One of them
is when the pigment in the resin(base) adsorbs OH groups onto
夽 This paper is modified version of the paper presented at the conference EuroCorr its surface thus leaving fewer groups available to react with iso-
2013 held in Estoril, Portugal, 1st to 5th September 2013. cyanate [4]. This means that the actual hardener to resin ratio may
∗ Corresponding author. Tel.: +44 16043213.
be considerably more than the stoichiometric ratio. Also during
E-mail addresses: ewa.papaj@hotmail.com (E.A. Papaj),
douglas.mills@northampton.ac.uk, Douglas@harrbridge.freeserve.co.uk application excessive hardener may be added to achieve a higher
(D.J. Mills), ssjamal@northampton.ac.uk (S.S. Jamali). gloss level [5] leading to a hardener to resin ratio that is lower than

http://dx.doi.org/10.1016/j.porgcoat.2014.08.013
0300-9440/© 2014 Elsevier B.V. All rights reserved.

Please cite this article in press as: E.A. Papaj, et al., Effect of hardener variation on protective properties of polyurethane coating, Prog.
Org. Coat. (2014), http://dx.doi.org/10.1016/j.porgcoat.2014.08.013
G Model
POC-3474; No. of Pages 5 ARTICLE IN PRESS
2 E.A. Papaj et al. / Progress in Organic Coatings xxx (2014) xxx–xxx

recommended. Environmental moisture can also react with iso- DC measurements were made with a Keithley Electrometer
cyanate groups resulting in an lower NCO/OH ratio together with, Model 610C using a two electrode system. Using the minimum
in some cases, structural defects [6]. DC current enables an accurate measurement to be made with no
The main aim of this work was to evaluate an acrylic significant destructive effect on the coatings being tested.
polyurethane coating with hardener/resin ratios of 0.75, 1.0 and Abrasion test was performed using the Taber abrasion tester
1.5. Protective properties of coatings were assessed in laboratory with CS-10 wheels and 500 g load. The adhesion strength of coated
immersion tests using Electrochemical Impedance Spectroscopy surfaces was measured using an automatic Defelsko PosiTest Pull-
(Impedance) and DC resistance. Also efforts have been made to off adhesion tester model AT-A at pull rate of 0.70 MPa/s with
find out the effect of hardener variation on mechanical proper- 20 mm diameter size of aluminium dollies. Recovery (wet) adhe-
ties, cross-linking and chemical structure of acrylic polyurethane sion measurements were made by removing the samples from
coatings. These aspects were investigated using Taber abrasion test, immersion, drying the surface and attaching the dolly to the coating
pull-off adhesion test, Dynamic Thermo-Mechanical Analysis and surface using cyanoacrylate super-glue. Each glued dolly was given
Fourier Transform Infrared Spectroscopy. a few minutes to completely set before the measurement was con-
ducted. Using steel substrates with 2 mm thickness and dolly size of
2. Experimental 20 mm diameter with the automated pull-off mechanism enabled
a reproducible measuring of adhesion strength with no distortion
2.1. Materials of substrate after measurement.
DMTA examination was performed using a Triton DMA model
Acrylic polyurethane varnish with three hardener ratios was Tritec 2000 with tension deformation mode at 1 Hz oscillation. The
prepared. Coatings components were: Setalux 1196 XX-60 (sup- displacement was 0.02 mm, the start temperature was 22 ◦ C and
plied by Nuplex) as resin, Tolonate HDT-LV2 (supplied by Perstorp) the end temperature was 100 ◦ C with the ramp rate of 2 ◦ C/min.
as hardener and ethyl acetate as thinner. No additional pigment or A SHIMADZU Fourier Transform Infrared Spectrophotometer
fillers were incorporated. FTIR-8400S spectrophotometer with a Smiths Detection DuraSam-
Mild steel Q-panels, 75 mm × 150 mm (purchased from Q-Lab), plIR II ATR attachment was used. Using the ATR attachment,
were used as metal substrates. Polypropylene sheets were used to IR spectrum was collected in reflectance mode and transmit-
obtain detached coatings. tance spectrum was collected with the wavelength range of
700–4500 cm−1 with resolution of 4.0 cm−1 .

2.2. Sample preparation

3. Results

The resin and hardener were mixed in three ratios: 0.75 (less
3.1. Electrochemical Impedance Spectroscopy
hardener than recommended), 1.0 (as normally recommended) and
1.5 (more hardener than recommended). The calculations used to
Coated samples were exposed to 3% sodium chloride contained
obtain each ratio can be found elsewhere [7]. To obtain an attached
in attached cells (two with each hardener to resin ratio). The hard-
uniform coating of two different thicknesses the resin/hardener
ener to resin ratio was: A, B: 0.75; C, D: 1.0 and E, F: 1.5. Thin coatings
mix was applied using the appropriate spreader bar onto steel Q-
are A, C and E and thick coatings are B, D, F. EIS measurements
panels. To get a detached coating, it was applied to polyethylene
were made periodically up to 1008 h. Results showing the effect of
sheets backed on a glass plate. All panels and sheets were degreased
time on the 0.1 Hz impedance value are shown in Fig. 1 (the value
with iso-propanol before coating application. After application, the
of impedance at 0.1 Hz is taken as the overall impedance of the
coatings were cured, first in a fume hood for 24 h (air drying), next
coated samples and allows for easy comparison with the DC resis-
at 50 ◦ C in oven for 24 h (forced drying). For EIS and DC resistance
tance.) Results show the coatings with hardener to resin ratio of
tests, short lengths of plastic tubes were affixed to the coated panels
1.5 have the highest impedance. Thick coatings have higher values
with silicone sealant to create two “cells” on each panel. The area
than thinner coatings and in general all Impedances decrease with
of each cell was 10.4 cm2 . Cells were filled with 3% sodium chloride
time.
and the exposure continued for up to 1008 h. The thickness of sam-
ples used was 75–90 ␮m for “thin” coatings and 100–130 ␮m for
3.2. DC resistance of coated samples
“thick” coatings. For the abrasion test, the samples had a specific
shape, like a truncated circle with diameter of 107 mm, to the width
The designation of the coated samples for DC resistance mea-
of 75 mm. The thickness of samples were ca. 80 ␮m for thin and ca.
surement was the same as those used in the EIS test. The results of
110 ␮m for thick coatings. For the wet adhesion, filter paper soaked
the DC measurements are shown in Fig. 2. After 1008 h the order
with 3% sodium chloride was placed onto the coated panel for 2 and
of the thin coating resistance is slightly different to that obtained
5 days. Two pieces of the filter paper were on each panel. The thick-
using EIS. The order of thick coating DC resistance follows very
ness of samples were ca. 75 ␮m for thin and ca. 115 ␮m for thick
much the same order as with EIS.
coatings. For DMTA examination, detached coatings, 60–70 ␮m,
were used. Each sample had a rectangular shape with dimensions
3.3. Abrasion test
of 5 mm × 11 mm. For FTIR analysis the detached coating did not
have a special shape or thickness. The reference spectra were col-
The weight of each sample was measured before and after the
lected using liquid resin and liquid hardener, which provide spectra
abrasion cycles with 0.0001 g accuracy and the Taber Wear Index
of unreacted OH and NCO groups.
was calculated according to Eq. (1):
Weight Loss in mg
2.3. Test methods Taber Wear Index = (1)
Number of Test Cycles
EIS measurements were performed using an ACM potentio- The lower the Taber Wear Index, the better the abrasion
stat/galvanostat/FRA model Gill AC. The standard three electrode resistance. The results for thin and thick coatings with different
arrangement was employed and readings were obtained from hardener ratios are shown in Fig. 3. The coating with hardener to
10 kHz to 0.03 Hz using a 150 mV applied AC signal. resin ratio of 1.5 has the best abrasion resistance. The coatings with

Please cite this article in press as: E.A. Papaj, et al., Effect of hardener variation on protective properties of polyurethane coating, Prog.
Org. Coat. (2014), http://dx.doi.org/10.1016/j.porgcoat.2014.08.013
G Model
POC-3474; No. of Pages 5 ARTICLE IN PRESS
E.A. Papaj et al. / Progress in Organic Coatings xxx (2014) xxx–xxx 3

Fig. 1. Low frequency (0.1 Hz) impedance obtained from EIS as a function of time for (a) thick and (b) thin polyurethane coatings.

(b)

Fig. 2. DC resistance as function of time for (a) thick and (b) thin polyurethane coatings.

0.75 hardener ratio show significantly less high abrasion resistance.

Thicker coatings appear to have slightly better abrasion resistance
than thinner coatings.

3.4. Adhesion test

The adhesion test measurements were performed after 2 and 5

days immersion. The repeatability of adhesion results was within
±20% and the averages are plotted in Figs. 4 and 5. Fig. 4 shows
the overall effect of hardener to resin ratio on adhesion, calculated
as an average for both coating thicknesses throughout the whole
duration of immersion.
The lowest adhesion was shown by the coating with hardener
to resin ratio of 1.0. Coatings with hardener ratio of 0.75 and 1.5
Fig. 4. Results of wet adhesion test for samples with varying hardener ratio.
showed similar values of adhesion, both considerably higher than
the coating with the correct ratio. The individual results (after
2 and 5 days) for thin and thick coatings with the same hard-
ener ratio are shown in Fig. 5. After 5 days both thin and thick

Fig. 3. Comparison of Taber Wear Index for polyurethane coatings with different Fig. 5. Wet adhesion of polyurethane coatings to steel with varying thickness after
hardener ratios. 2 and 5 days.

Please cite this article in press as: E.A. Papaj, et al., Effect of hardener variation on protective properties of polyurethane coating, Prog.
Org. Coat. (2014), http://dx.doi.org/10.1016/j.porgcoat.2014.08.013
G Model
POC-3474; No. of Pages 5 ARTICLE IN PRESS
4 E.A. Papaj et al. / Progress in Organic Coatings xxx (2014) xxx–xxx

Fig. 6. FTIR spectrum of coatings with different hardener/resin ratios compared with (a) pure liquid resin and (b) pure liquid hardener.

coatings show considerably reduced adhesion compared to the 2 the hardener/resin ratio as shown in Fig. 3, can be explained by
days immersion results. Thicker coatings have a better adhesion increased storage modulus (toughness) and elasticity brought
than thinner coatings and this is particularly marked after 5 days. about by the improved polymeric network [11]. The fact that the
abrasion resistance of the thick coating is slightly higher than that
3.5. Fourier Transform Infrared Spectroscopy of the thin coating, may be an artefact of the Taber method. More
tests are required to examine the significance of variation in abra-
Individual functional groups of polyurethane coatings were sion resistance. Regarding the effect of hardener ratio on adhesion
studied using FTIR spectroscopy as shown in Fig. 6. The char- shown in Figs. 4 and 5, it seems likely that non-reacted functional
acteristic functional groups of polyurethane coatings can be groups present in both the 1.5 and 0.75 hardener to resin ratio
found in following regions of the FTIR spectrum: 3330 cm−1 : samples might increase the adhesion to substrate (e.g. through the
N H; 2960–2963 cm−1 : C H; 2865–2873 cm−1 : C H; 2280 cm−1 : electrostatic bonds) [12,13]. Another possible explanation is that
N C O; 1721–1738 cm−1 : C O; 1640 cm−1 : NH C( O) O. the imperfect coating (0.75 or 1.5) is more viscous than the coating
Results show that the coating with hardener to resin ratio of 0.75 with the stoichiometric hardener to resin ratio of 1 and therefore
has a higher OH peak indicating more unreacted OH groups. The it deforms more easily [14]. So when performing the pull-off test
coating with hardener ratio of 1.5 has a higher NCO peak. it deforms before it completely breaks. With similar pull-rate it
detaches later than the stoichiometric composition resulting in
3.6. Dynamic thermo-mechanical analysis higher adhesion value. The DMTA measurements provide values
of Tg . This is connected with the cross link density of coating [15].
The glass transition temperature (Tg ) was measured using DMTA Hence it appears that the coating with hardener to resin ratio
technique as the maximum of damping peak in the tan ı curve. of 1.5 has significantly higher cross link density than that with
Measurement was repeated three times and the average is reported 1.0. This relates to the impedance resistance results where again
as Tg in Table 1. The higher the hardener to resin ratio, the higher the samples with higher hardener to resin ratio had the higher
the Tg . There is a particularly noticeable increase in Tg between a impedance/resistance.
hardener ratio of 1 and of 1.5. The presence of more unreacted groups in the 0.75 and 1.5
samples is shown in the FTIR spectroscopy results, Fig. 6. The
4. Discussion 3600–3500 cm−1 peak is a characteristic for the resin and denotes
the presence of hydroxyl group (O H). The largest OH peak was
Regarding the results of the electrochemical tests shown in observed for the coating with hardener ratio of 0.75 indicating extra
Figs. 1 and 2, the gradual decrease in impedance/resistance with unreacted OH groups in the structure of this coating. The 2256 cm−1
time is typical of the behaviour of anti-corrosive coatings on steel peak is attributed to isocyanate group (N C O), which is charac-
[8] as more water enters the coating and facilitates the movement teristic of the hardener. This peak is most significant in the case
of ions. The fact that thicker coatings have higher resistances is of the coating with hardener ratio of 1.5, which proves there are
also explicable because of the overall more difficult path to ions unreacted NCO groups in this structure.
that thicker coatings offer and their lower water absorption [9].
The fact that the best resistance is offered by the 1.5 hardener to
5. Conclusions
resin ratio may be a combination of higher cross linking (as shown
by DMTA test) and also higher adhesion (the 0.75 hardener ratio
Overall this work has shown that varying the hardener to resin
also offers better adhesion; but not such high cross linking) [10].
ratio has quite a major effect on both electrochemical, (protec-
The fact that the abrasion resistance is increased by increasing
tive) properties as well as mechanical properties of polyurethane
coatings. Bearing in mind there are number of scenarios where a
Table 1 different ratio to the correct one may be present in the final dried
The glass transition temperature for coatings with various hardener to resin ratios
measured by DMTA.
film, the user and manufacturer of these coatings should take this
into account.
Hardener ratio of Tg [◦ C] Acrylic polyurethane coatings with a higher hardener/resin ratio
0.75 58.9 showed higher electrical (ionic) resistance as revealed by EIS and
1.0 60.7 DC resistance measurements. Therefore they may be expected to
1.5 67.5
offer better protection. The probable reason for this, as shown

Please cite this article in press as: E.A. Papaj, et al., Effect of hardener variation on protective properties of polyurethane coating, Prog.
Org. Coat. (2014), http://dx.doi.org/10.1016/j.porgcoat.2014.08.013
G Model
POC-3474; No. of Pages 5 ARTICLE IN PRESS
E.A. Papaj et al. / Progress in Organic Coatings xxx (2014) xxx–xxx 5

by the DMTA test, could be the higher crosslink density of the Acknowledgements
coating with higher hardener/resin ratio. Also acrylic polyurethane
coatings with a higher hardener ratio showed better abrasion The authors would like to thank Rachel Garfoot, Head of the
resistance. This appears to depend to a small extent on the thick- Leather Sellers Centre at the University of Northampton for making
ness of the coating. Acrylic polyurethane coatings with the lower available laboratories (and the equipment therein) where the bulk
hardener ratio have fewer urethane groups than coatings with of the work was carried out. Also we would like to acknowledge the
the higher hardener ratio, which has been proven by use of the Erasmus programme for supporting one of the authors (EP) during
FTIR spectroscopy method. This may help explain why the adhe- her time in Northampton. And we would also like to thank Paul
sion values of acrylic polyurethane coatings with correct hardener Alexander of Pronto Paints for taking an interest in the project and
ratios are the lowest relating to the fact that all the NCO and making useful suggestions. Also thanks are expressed to Nuplex
OH groups this coating have been used up in internal bonding. and Perstorp for provision of materials.
There is also an indication that the thin coating has poorer adhe-
sion than the thicker coatings although further work is needed to References
prove this.
The work also shows that there is a need for greater aware- [1] W. Morgans, Outlines of Paint Technology, Halsted Press, New York, 1990
(Chapter 11).
ness by paint manufacturers, applicators and end users, of the effect [2] N. Tuck, M. Bardour, J. Clarke, D. Fone, Two component isocyanate cured acrylics
of changing the hardener/base ratio on properties. It appears that applications, in: P. Oldering, P. Lam (Eds.), Waterborne and Solvent Based
using a binder to hardener ratio different to that recommended by Acrylics and Their End User Applications, John Wiley & Sons, Inc., London, 1996,
p. 289.
the manufacturer (particularly increasing the ratio beyond manu- [3] M.B. Dromgool, Advantages of using a polyurethane topcoat, J. Protect. Coat.
facturer’s recommendations) is not necessarily deleterious and in Linings (1999) 11–13, January.
fact may in certain cases positively influence the coating proper- [4] M.S. Ceh, Urethane coatings, in: L.J. Calbo (Ed.), Handbook of Coatings Additives,
vol. 2, Marcel Dekker Inc., 1992, p. 263.
ties. However it also bears the risk of losing any warranty from the
[5] U. Meier-Westhues, Polyurethanes – Coatings, Adhesives and Sealants, Vin-
paint manufacturer. centz Network, Hanower, 2007.
Finally Electrochemical Impedance Spectroscopy and the DC [6] C.H. Hare, A review of polyurethanes: formulation variables and their effects
resistance method have proved themselves to be good methods on performance, J. Protect. Coat. Linings (2000) 34–44, November.
[7] P. Ardaud, J.M. Bernard, C. Varron (Eds.), Water Borne and Solvent Based Sur-
for investigating how the resistance of coatings depends on both face Coating Resins and Their Applications. Vol. III: Polyurethanes, Wiley/SITA,
the nature of the coating and time. London, UK, 1998.
[8] S.W. Tait, An Introduction to Electrochemical Corrosion Testing for Practicing
Engineering and Scientists, Pair o Docs Publications, Racine, WI, 1994.
[9] D. Greenfield, J.D. Scantlebury, The protective action of organic coatings on
Accolade steel: a review, J. Corros. Sci. Eng. 3 (2000) (paper 5).
[10] F. Levine, J. Escarsega, J. La Scala, Effect of isocyanate to hydroxyl index on the
properties of clear polyurethane films, Prog. Org. Coat. 74 (2012) 572–581.
One of the authors (DM) would like to acknowledge the oppor- [11] D.K. Chattopadhyay, K.V.S.N. Raju, Structural engineering of polyurethane
tunity that Gordon Bierwagen gave to him which was to come coatings for high performance applications, Prog. Polym. Sci. 32 (2007)
and work in the Department of Polymers and Coatings at NDSU 352–418.
[12] C.H. Hare, I. Adhesion, J. Protect. Coat. Linings (1996) 79–95, May.
in Fargo as a visiting scientist for over two years in the mid-90s. [13] C.H. Hare, Adhesion: Part 2, J. Protect. Coat. Linings (1996) 79–95, July.
This period produced not only a large number of publications but [14] Z. Sun, K.-T. Wan, D.A. Dillard, A theoretical and numerical study of thin film
also a considerable expansion of his research interests. That then delamination using the pull-off test, Int. J. Solids Struct. 41 (2004) 717–730.
[15] R. Mafi, S.M. Mirabedini, M.M. Attar, S. Moradian, Cure characterization of epoxy
led on to the eventual setting up in Northampton of his own small
and polyester clear powder coatings using Differential Scanning Calorimetry
group researching into anti-corrosive coatings and development of (DSC) and Dynamic Mechanical Thermal Analysis (DMTA), Prog. Org. Coat. 54
improved electrochemical methods. This paper is an outcome. (2005) 164–169.

Please cite this article in press as: E.A. Papaj, et al., Effect of hardener variation on protective properties of polyurethane coating, Prog.
Org. Coat. (2014), http://dx.doi.org/10.1016/j.porgcoat.2014.08.013