Journal of Dentistry (2005) 33, 585–591

www.intl.elsevierhealth.com/journals/jden

Thermal conductivity through various restorative

lining materials
Peter AG Littlea, David J Wooda, Nigel L Bubba, Stuart A Maskilla,
Lawrence H Mairb, Callum C Youngsonb,*
a
Division of Restorative Dentistry, Leeds Dental Institute, Leeds LS2 9LU, UK
b
Department of Clinical Dental Science, Unit of Restorative Dentistry, Liverpool University Dental School,
Liverpool L3 5PS, UK

Received 10 February 2004; received in revised form 21 December 2004; accepted 24 December 2004

KEYWORDS Summary Objectives. Traditional teaching has advocated the placement of

Dental materials;
Thermal conductivity;
Cavity lining
insulating materials under restorations to protect against ‘thermal shock’. Often
this involves placement of a thick zinc-oxide eugenol cement lining. The purpose of
this in vitro study was to assess the heat transfer through four lining materials and
dentine and to relate these findings to the temperature exposures that may be
experienced in the oral environment.
Methods. Two ‘traditional’ linings, zinc oxide eugenol cement (ZNOE) and calcium
hydroxide paste (CAOH), and two more contemporary materials, a resin modified
glass polyalkenoate (RMGP) and dentine bonding system (DBS) were studied in
addition to dentine itself. Seven samples of each material were tested by utilising a
Lees’ Disc apparatus. Incorporation of thermocouples allowed heat transfer through
the sample to be recorded. Thermal conductivity (J/m/s/8C) was calculated for
individual samples and a mean value obtained for each material. One way ANOVA and
post hoc Tukey’s tests were applied to the data obtained.
Results. In ascending order of thermal conductivity, DBS!CAOH!RMGP!
DENTINE!ZNOE. Statistical analysis revealed that the thermal conductivity of DBS
was significantly less than RMGP, DENTINE and ZNOE (p!0.05); CAOH was
significantly less than DENTINE and ZNOE (p!0.05); RMGP was significantly higher
than DBS and significantly less than DENTINE and ZNOE (p!0.05).
Conclusions. Thermal insulation is not a major consideration in cavity lining and
therefore other criteria for selection of a lining material should be applied.
q 2005 Elsevier Ltd. All rights reserved.

Introduction

* Corresponding author. Tel.: C44 0151 706 5230; fax: C44

Traditional teaching has advocated the use of lining
0151 706 5845. materials to afford protection to the pulp and
E-mail address: c.c.youngson@liv.ac.uk (C.C. Youngson). insulate the pulp from the extremes of thermal
0300-5712/$ - see front matter q 2005 Elsevier Ltd. All rights reserved.
doi:10.1016/j.jdent.2004.12.005
586
stimuli particularly after restorative procedures.
The best insulating material available has been
quoted as being dentine itself.1 Rationales behind
lining placement, with consideration of individual
materials and their possible applications, have been
reviewed2,3 but little evidence-based explicit gui-
dance currently exists. The current rationale
behind the placement of a lining would appear to
be to attempt to reduce dentinal fluid flow and
bacterial ingress, as these appear to be the cause of
much of the post-operative sensitivity experienced
by patients.4
Zinc-oxide eugenol has been the classic dental
base material for many generations of dentists who
were encouraged to place thick linings to ensure
good thermal insulation. It has been for many years
considered to be the ideal lining material.5 It has
insulating properties6 and proven antibacterial
properties.7 Calcium hydroxide has been used as a
lining and sub-lining material for several decades.8
Its therapeutic effects on the exposed human pulp9
and ability to promote dentine bridging have been
suggested as the chief reasons for using this
material as a lining.10 Resin modified glass polyalk-
enoates (RMGP) may bond to dentine via mechan-
ical,11 micromechanical12 or chemical adhesion.13
This group of materials have demonstrated good
dentinal sealing properties in vitro.14,15 Dentine
bonding systems, which seal the tubules from
external stimuli, e.g. hot and cold, should prevent
hydrodynamic fluid shifts, thereby stopping the
stimulation of the nerve fibres and the resulting
pain. This has been demonstrated clinically16 in an
investigation that also cast doubt on the need for a
thick insulating layer to ‘protect’ the dentine and
prevent post-operative sensitivity.
The purpose of the current study is to assess the
thermal transfer of dentine and commonly used
lining materials via conductivity experiments.
These values can then be considered in relation to
the clinical situation that involves only transient
thermal extremes17,18,19 to determine whether a
bulk thickness of lining material is necessary.
Materials and methods
The following materials were tested for thermal
conductivity.
1. Zinc oxide eugenol cement. (ZNOE), Kalzinol,
Dentsply DeTrey, Milford, DE, US
2. Resin modified glass polyalkenoate. (RMGP)
Vitrebond, 3MEspe, St Paul, MN, US
3. Dentine bonding system. (DBS) Scotchbond 1,
3MEspe, St Paul, MN, US
P.A.G. Little et al.
4. Calcium hydroxide lining, chemical setting.
(CAOH) Dycal, Dentsply DeTrey, Milford, DE, US
5. Dentine (extracted human first permanent molar
teeth)
Both the dentine bonding system and the resin
modified glass polyalkenoate were light cured using
a Visilux 2 VLC, (Model 5520 CP, 3MEspe, St Paul,
MN, US) that had its light output confirmed as
satisfactory prior to use.
Sample preparation
Copper tubing was cut into 1 mm thick rings to act
as moulds. Seven discs were produced for each of
the four lining materials tested. Materials were
handled according to manufacturers’ instructions
and placed within the copper rings on a clean
microscope slide. A second slide was then used from
above to compress the materials into disc form and
ensure smooth parallel disc walls. Excess material
was removed by scalpel.
ZNOE and CAOH discs were left to harden for
10 min between the slides before they were
removed. RMGIC was cured for 60 s from one side
only. DBS was placed into the copper rings and after
compressing with the second slide, was light cured
from one side (for 5 min). All materials were stored
for at least 24 h after fabrication before testing to
ensure that no exotherm associated with their
setting affected subsequent results.
Samples of dentine were obtained from unrest-
ored extracted human first permanent molar teeth.
The discs were first sectioned from just above the
pulp chamber and then approximately 2 mm above
this level. The sections were then ground smooth on
either side using a polishing disc (MK 2A, Engis Ltd,
Maidstone, Kent, UK) with water coolant and an
attempt was made to maintain parallel sides by
continually checking the thickness using a
micrometer. At this level of section the dentine
was generally circular. The outer shell of enamel
was removed using the polishing disc.
Three different measurements of the diameter
were made of each sample and a mean value
obtained. Three different thickness measurements
were also made of each sample and a mean value
was calculated for each.
The measurement of thermal gradients and
conductivity involved the construction of a Lees’
Disc20 (Fig. 1). A cylindrical steam chamber had at
one end, a circular brass disc approximately 20 mm
in diameter and 3 mm thick. A second disc (made
from 99.99% copper), of similar size was fabricated.
Between these two discs, the sample materials
Thermal conductivity of cavity liners
Figure 1 Diagram of steam chamber with sample and
copper disc.
(12 mm!1 mm) or dentine discs were placed. Both
the brass end of the steam chest and the free
copper disc were flattened and highly polished. The
mass of the free copper disc was found to be
12.61 g.
Small holes were drilled into the sides of the
brass and copper discs which would accommodate
two-welded tip, PTFE insulated K-type thermo-
couples (R.S. Components Ltd, Corby, Northants,
UK), one in each disc, to record temperature
differentials. The thermocouples were capable of
measuring from K50 to 200 (G3) 8C.
The steam chamber apparatus was assembled
vertically with the polished end at the top. This
surface was lightly brushed with a thin layer of
petroleum jelly, as was the copper disc and the
sample disc of material, to ensure a good contact
for heat transfer. The sample disc was then situated
on top of the steam chamber and the copper disc
(with thermocouple inserted) placed on top of the
sample.
The thermocouples were connected via a Digi-
tron Datalogger (version 1.14, Digitron Instrumen-
tation Ltd, Hertford, UK) to a Psion Organiser II,
model LZ64 (Kuma Computers Ltd, Berkshire, UK).
Data was transferred to a personal computer using a
Loglink Analysis programme (Digitron Instrumenta-
tion Ltd) and entered onto a spreadsheet (Microsoft
Excel, Microsoft, USA) to allow analysis. All pro-
cedures took place in a closed-room environment
with temperature controlled at 21G1 8C.
When the steam chamber had reached its
equilibrium, temperatures were logged every five
seconds from when the copper and brass discs were
initially brought together, separated by the sample
disc, until a steady state was reached. The sample
was then removed from between the two discs and
these were placed together to allow a second
steady state to be reached; meanwhile the sample
587
disc was kept warm by placing it on the underside of
the steam chamber. Once the second steady state
was reached, the free copper disc was removed
from the top of the steam chamber and allowed to
cool with the sample placed on top of it. This was
effectively held in mid-air by the thermocouple
wires to ensure no interference heat loss occurred;
the temperatures were still being recorded every
5 s. Once the free copper disc plus sample
temperature had fallen at least 15/20 8C below
the original steady state condition the data record-
ing was halted.
Thermal conductivity calculations
The rate of flow of heat from the hotter face to the
colder face (dQ/dt) equals the temperature gradi-
ent (dq/dx), multiplied by the cross sectional area
(A), multiplied by the thermal conductivity value
(k), i.e.
dQ =dt Z kA dq=dx
In the case of the Lees’ Disc, the above equation
can be re-written as:
dQ =dtZ kAðq2 K q1Þ=x
From the equation above, it was important to
obtain a value for q2Kq1, (the temperature of the
steam chamber at steady state minus the tempera-
ture of the copper disc at steady state) for each
sample disc. The steady state temperatures were
calculated using an average of twenty correspond-
ing values before the sample disc was removed.
The free copper disc was removed from the top
of the steam chest and the sample placed simul-
taneously upon it; both were then suspended in air
as they cooled together. The temperature was
recorded every five seconds until it was at least
15 8C below the steady state condition to obtain the
thermal gradient (a/b). The disc had, at this point,
been losing heat at the same rate as it did in the
steady state conditions and a graph of temperature
against time (a/b), could be plotted from the data
recorded as shown in Fig. 2. Gradients (a/b) were
obtained using linear regression.
The rate at which the disc lost heat to its
surroundings when the temperature was at the
steady state is given by:
mca=b
where: m equals the mass of the copper disc, (in
this case 12.61x10K3 kg); c equals the specific heat
capacity of the copper disc, 385 J/kg/8C;21 and a/b
is the gradient of heat loss, obtained from the data.
Properties of copper used in the apparatus are:
588 P.A.G. Little et al.

Results

A typical graphical representation of the data, in

Figure 2 Temperature recorded by thermocouple
inserted into copper disc for ZNOE sample.
density 8.93!10 3 kg/m 3 and conductivity
385 J/m/s/8C.21 The conditions under which the
copper disc was losing heat were the same as those
at the steady state, therefore,
kAðq2 K q1Þ=x Z mca=b
From the above equation, the value of k, the
thermal conductivity co-efficient, could be calcu-
lated. Constant values could then be substituted in
for all samples:
k Z x !12:61 !10K3 !385 !a=b=pr2 !ðq2 K q1Þ
where:
sample thicknessZx
mass (m) of copper discZ12.61!10-3 kg
specific heat capacity (c) of copperZ
385 J/kg/8C
Area (A) of sample discZpr2
Using this equation an average value for K and its
associated standard deviation were obtained for
each group.
During testing two samples (1 each from the DBS
and DENTINE groups) fractured. All data from these
samples was excluded from analysis.
this case for ZNOE, is shown in Fig. 2. The first
region corresponds to when the copper and brass
discs were initially brought together, separated by
the sample disc, until a steady state was reached.
The increase in temperature then corresponds to
when the sample was then removed from between
the two discs and these were placed together to
allow a second steady state to be reached. The third
region corresponds to when the free copper disc
was removed from the top of the steam chamber
and allowed to cool with the sample placed on top
of it. The gradient (a/b) was recorded in this region.
Using Minitab software, statistical analysis
involved the calculation of the standard deviations
and the standard errors of the means with 95%
confidence intervals (Table 1). In ascending order of
thermal conductivity, DBS!CAOH!RMGP!DEN-
TINE!ZNOE
A one way analysis of variance (ANOVA) using
Tukey’s pairwise comparisons was then performed.
This analysis revealed statistically significant
differences for thermal conductivity:
DBS was significantly different to RMGP, DENTINE
and ZNOE (p!0.05).
CAOH was significantly different to DENTINE and
ZNOE (p!0.05).
RMGP was significantly different to DBS, DENTINE
and ZNOE (p!0.05).
DENTINE was significantly different to DBS, CAOH
and RMGP (p!0.05).
ZNOE was significantly different to DBS, CAOH
and RMGP (p!0.05).
Discussion
Thermal conductivity has been defined as ‘The rate
of flow of heat per unit area per unit temperature
gradient when heat flow is under steady state
conditions’.21 However, it is not the steady state
condition at 37 8C that is of concern in the mouth

Table 1 Statistical analysis of data.

Level N Mean Std. Dev. 95%CI
CAOH 7 0.41 0.07 0.36–0.47
DBS 6 0.29 0.05 0.26–0.33
RMGP 7 0.58 0.12 0.49–0.66
ZNOE 7 0.93 0.19 0.80–1.07
DENTINE 6 0.88 0.19 0.73–1.03
Thermal conductivity of cavity liners
but the conduction of temperature extremes to the
pulp. Under these conditions thermal diffusivity
may be considered more relevant. The values for
thermal diffusivity can be calculated from the
equation:
h Z k=cr
where:
h is the thermal diffusivity; k is the thermal
conductivity; c is the specific heat capacity, and r is
the density of the material.
When using composite restorations another
factor, the exotherm associated with the setting
material, should be considered. Using thermo-
graphic imaging techniques, one group of workers
have demonstrated mean temperature rises of
5.4 8C in vivo with a peak exotherm at 30 s after
commencing curing in Class III or Class IV restor-
ations22. As these procedures are often performed
on patients under local anaesthesia pulpal trauma
may not become immediately apparent. The inves-
tigation of this was, however, beyond the remit of
the current investigation.
It is hoped that the data obtained in this study
will be of use in future research in determining
thermal diffusivities for these materials. Notwith-
standing this limitation the current data can be
applied, indirectly, to the clinical situation.
The traditional belief that linings should prevent
thermal shock and replace large quantities of lost
dentine appear to a certain degree to be without
scientific basis as demonstrated by the scenarios
outlined below, using data for dentine generated in
this study.
It has been stated that the best defence to a
pulpal insult is the remaining thickness of dentine,
and the more that remains the better.1 Dentine
does not, on the basis of the current findings,
appear to be a very efficient thermal insulator with
a mean conductivity of 0.88 J/m/s/8C (in keeping
with data reported early in the 1950s22) and it had a
significantly higher conductivity value than DBS,
CAOH or RMGP.
If hot liquid is taken into the oral cavity at 65 8C,
it will arrive on the tooth surfaces at a maximum of
50/55 8C after the relatively cooling effects of lips
and mucosa.18,19 It could be assumed that, if this
temperature reaches the top of a deep unlined
amalgam with a remaining dentine thickness of
1 mm, all the 55 8C is conducted through the
amalgam to the amalgam dentine interface. A
1 mm thick layer of dentine (similar to the test
specimen) would then be challenged by a tempera-
ture of 55 8C. This is clearly lower than the steam
chamber challenge of 99 8C from which the thermal
589
gradient of heat loss for dentine (mean value 0.18)
was obtained and therefore we might expect a
lower thermal gradient. If the gradient of heat loss
determined by the conductivity of dentine is 0.18,
then:
a=b Z 0:18
One group of investigators23 have reported that an
11.2 8C rise in temperature at the dentinopulpal
junction was acceptable and that pain was felt at
temperatures approaching 45 8C. Others24 have
concluded that a safety limit of 6.1 8C above
normal pulpal temperature was applicable. Sub-
stitution in the above equation of a temperature
rise of 6 8C will give the time required for this
elevation to take place (i.e. the time taken to
raise the temperature at the dentinopulpal inter-
face from 36 to 42 8C).
b Z 6=0:18 Z 33 s
The gradient for dentine was obtained using a
heating temperature of approximately 99 8C, an
unrealistic figure in the oral cavity, therefore the
time taken to raise the temperature by this much
could be even longer than this. This value also
does not allow for heat loss within the amalgam
and lateral heat diffusion that would further
increase the time required to increase the
temperature at the dentinopulpal surface. The
above equation illustrates the point that consider-
able and constant temperature elevations applied
for long periods are necessary to induce possible
thermal damage in pulpal tissue. Within the mouth
this would be probably be unbearable, due to
discomfort being experienced from the oral
mucosa, and impossible to achieve, due to con-
duction of the heat from the mouth through the
mucosal blood supply. It has been shown that
intra-oral temperature extremes are transient
during normal function.17–19 If we assume using
data from intra-oral studies that the highest intra-
oral temperature is present for 3 seconds then:
a=5 Z 0:18
a Z 0:9 8C
i.e. the temperature at the dentinopulpal junction
will rise by 0.9 8C if 1 mm of dentine is present.
Thermal insulating efficiency is a function of the
material’s thickness divided by the square root of
its diffusivity25 and insulating efficiency is directly
related to the thickness of a material. Therefore we
might expect the time taken to raise the tempera-
ture by 6 8C to be 16.5 s (33/2) if the remaining
dentine thickness is only 0.5 mm. This is still,
however, unlikely to be of significance during
590
normal function. As this work, utilising relatively
high temperatures, has failed to establish a need
for thermal insulation it appears that there is no
need to place an ‘insulating layer’ to protect the
pulp. ZNOE will, however, provide this function
should a clinician wish to replace lost dentine or
gain the other therapeutic benefits of it. Dycal
(setting CaOH paste) has a conductivity that is
significantly lower than ZNOE and so may be used in
thinner sections if the clinician feels that insulation
is important.
This investigation does not support the need for
thick linings under amalgam to prevent thermal
shock26 but the literature does support the need to
seal dentine after rotary instrumentation27. As
RMGP has this property15, and also had a statisti-
cally significantly lower thermal conductivity value
than either dentine or ZNOE, this would appear to
be an appropriate material for cavity lining. RMGP
also has three major advantages over ZNOE in that
it chemically bonds to the dentine, providing an
effective tubule seal, thereby preventing the
hydrodynamic fluid flows that might be associated
with thermal or chemical challenges. The second
advantage is that it can release a degree of fluoride
into areas were secondary caries could be a
problem. Thirdly it is ‘command set’ and therefore
easier to use.
The DBS tested in this study had the lowest
recorded mean conductivity value and was stat-
istically significantly lower than RMGP, ZNOE and
DENTINE. This material also has the ability, like
RMGP, to seal the tubules and therefore prevent
fluid evaporation and irritants from entering the
dentine. It is however used in very thin section
and so could be considered more of a cavity
sealer than a liner (which implies a degree of
thickness).
Conclusions
The two adhesive lining materials tested (RMGP and
DBS) that have the ability to seal the dentinal
tubules also had the lowest thermal conductivity
values.
The thermal conduction of dentine, as measured
in this study, suggests that only extreme tempera-
tures applied for long times would be harmful to the
pulp. As these are unlikely in vivo the insulating
property of a cavity lining material is not of great
significance.
In purely thermal terms DBS and RMGP are the
most efficient thermal insulators.
P.A.G. Little et al.
Acknowledgements
The authors would like to thank Ian Smith for
technical assistance during this study.
References
1. Stanley HR. Human responses to restorative procedures..
Gainesville: Storter Printing Co, Inc.; 1981.
2. Bartlett DW. A rationale for use of bases with amalgam
restorations. Dental Update 1992;19:413–7.
3. Hilton TJ. Cavity sealers. Liners and Bases: Current
philosophies and indications for use. Operative Dentistry
1996;21:134–46.
4. Brännström M. The cause of postrestorative sensitivity and
its prevention. Journal of Endodontics 1986;12:475–81.
5. Shroff FR. Effects of filling materials on dental pulp; an
histological experimental study with special reference to
synthetic porcelain. Pt 2. New Zealand Dental Journal 1947;
43:35–58.
6. Tibbetts VR, Schnell RJ, Swartz ML, Phillips RW. Thermal
diffusion through amalgam and cement base: comparison of
in vitro and in vivo measurements. Journal of Dental
Research 1976;55:441–51.
7. Markowitz K, Moynihan M, Lui M, Kim S. Biologic properties of
eugenol and zinc oxide eugenol: A clinically orientated
review. Oral Surgery Oral Medicine Oral Pathology 1992;55:
402–7.
8. Zander HA. Reaction of the pulp to calcium hydroxide.
Journal of Dental Research 1939;18:373–9.
9. Hebling J, Giro EM, Costa CA. Biocompatibility of an adhesive
system applied to exposed human dental pulp. Journal of
Endodontics 1999;25:676–82.
10. Foreman PC, Barnes IE. A review of calcium hydroxide.
International Endodontic Journal 1990;23:283–97.
11. Lin A, McIntyre NS, Davidson RD. Studies on the adhesion of
glass ionomer cements to dentine. Journal of Dental
Research 1992;71:1836–41.
12. Glasspoole EA, Erickson RL, Davidson CL. Effect of surface
treatments on the bond strength of glass ionomers to
enamel. Dental Materials 2002;18:454–62.
13. Titley KC, Smith DC, Chernecky R. SEM observations of the
reactions of the components of a light-activated glass
polyalkenoate (ionomer) cement on bovine dentine.
Journal of Dentistry 1996;24:411–6.
14. Youngson CC, Holguin SM. Early in vitro marginal microleak-
age associated with different lining materials under Class 2
amalgam restorations. European Journal of Prosthodontics
and Restorative Dentistry 1992;1:73–7.
15. Chuang SF, Jin YT, Lin TS, Chang CH, Garcia-Godoy F. Effects
of lining materials on microleakage and internal voids of
Class II resin-based composite restorations. American
Journal of Dentistry 2003;16:84–90.
16. Unemori M, Matsuya Y, Akashi A, Goto Y, Akamine A.
Composite resin restoration and postoperative sensitivity:
clinical follow-up in an undergraduate program. Journal of
Dentistry 2001;29:7–13.
17. Michailesco PM, Marciano J, Grieve AR, Abadie MJ:. An in
vivo recording of variations in oral temperature during
meals: a pilot study. Journal of Prosthetic Dentistry 1995;
73:214–8.
18. Youngson CC, Barclay CW. A pilot study of intraoral
temperature changes. Clinical Oral Investigations 2000;4:
183–9.
Thermal conductivity of cavity liners
19. Mair LH. Surface permeability and degradation of dental
composites resulting from oral temperature changes. Dental
Materials 1989;5:247–55.
20. Muncaster R. Advanced level physics. 2nd ed. Cheltenham:
Stanley Thornes Publishing Ltd; 1987.
21. Tennent RM, editor. Science Data Book. Edinburgh: Oliver
and Boyd; 1997.
22. Hussey DL, Biagioni PA, Lamey PJ. Thermographic measure-
ment of temperature change during resin composite
polymerization in vivo. Journal of Dentistry 1995;23:267–71.
23. Lisanti VF, Zander HA:. Thermal conductivity of dentine.
Journal of Dental Research 1950;29:493–7.
591
24. Baldissara P, Catapano S, Scotti R. Clinical and histological
evaluation of thermal injury thresholds in human teeth: a
preliminary study. Journal of Oral Rehabilitation 1997;24:
791–801.
25. Zach L, Cohen G. Pulp response to externally applied heat.
Oral Surgery Oral Medicine Oral Pathology 1965;19:515.
26. Braden M. Heat conduction in teeth and the effects of lining
materials. Journal of Dental Research 1964;43:315–22.
27. Bergenholtz G. Iatrogenic injury to the pulp in dental
procedures: Aspects of pathogenesis, management and
preventive measures. International Dental Journal 1991;
41:99–110.