The effect of impression volume and

double-arch trays on the registration of

maximum intercuspation
Sara M. Hahn,a Philip L. Millstein, DMD, MS,b Taru H.
Kinnunen, MA, PhD,c and Robert F. Wright, DDSd
Harvard School of Dental Medicine, Boston, Mass
Statement of problem. The type of double-arch trays used may affect occlusion.

Purpose. The purpose of this study was to determine what effect, if any, double-arch tray design and impression ma-
terial volume had on the registration of maximum intercuspation (MI).

Material and methods. Quadrant impressions were made on articulated fracture-resistant dental casts mounted in
maximum intercuspation occlusion. Three types of sideless double-arch impression trays were used: First Bite with
nylon webbing, Sultan’s 3-Way with double crosshatch webbing, and Premium’s 3-in-1 Tray with single crosshatch
webbing. Vinyl polysiloxane impression material (Aquasil Ultra Rigid Fast Set) was distributed at 2 different volumes
(5.4 ml and 8.3 ml), and 60 impressions were made (n=10). A weight of 1.2 kg was placed on the upper arm of an
Artex articulator, ensuring complete closure. The impressions were allowed to polymerize for 5 minutes. After polym-
erization, specimens were placed on a light box, and a camera set at a fixed distance was used to capture the light
transmission that was projected through the impression material. The camera transferred the information to an image
analysis program (ImageJ). This system allowed the different amounts of light projected through the impression to be
translated into a gray scale value (GSV), which was assigned a thickness value, in millimeters, of a specified occlusal
contact area. To assess reliability of the experimental design, 10 control impressions were made by directly applying
impression material onto the typodont. These were analyzed in the same manner as the impressions made with trays.
A 2-way ANOVA comparing volume by tray type was used (α=.05). This was followed by a Tukey HSD test.

Results. There was no main effect for volume of impression material (P=.71). Tray type was significantly different
(P<.001). Impressions made with Sultan trays were significantly less accurate than impressions made with First Bite
or Premium impression trays. The Premium tray type had the highest mean (SD) GSV, 179.8 (8.1), significantly higher
than First Bite (164.7 (19.8); P<.001) or Sultan (82.8 (7.4); P<.001) trays.

Conclusions. This in vitro study demonstrated that certain trays hinder closure into MI, which may negatively affect
the accuracy of record making. (J Prosthet Dent 2009;102:362-367)

Clinical Implications
Not all sideless double-arch impression trays are created equal.
Premium’s 3-in-1 tray and Dentsply’s First Bite trays provided a
more accurate registration of the occlusal position than Sultan
trays. The volume of impression material dispensed for a dou-
ble-arch tray has no effect on the occlusal relationship.

Funded by the Office of Enrichment Programs, Harvard Medical School, as well as the Prosthodontics Research and Scholarship
Fund and the Department of Restorative Dentistry and Biomaterial Sciences, Harvard School of Dental Medicine.

a
DMD student.
b
Lecturer, Restorative Dentistry and Biomaterials Sciences; private practice, Cambridge, Mass.
c
Assistant Professor, Department of Oral Health Policy and Epidemiology.
d
Associate Professor, Restorative Dentistry and Biomaterials Sciences; Director, Advanced Graduate Prosthodontics.
The Journal of Prosthetic Dentistry Hahn et al
December 2009
Dual-arch impression trays are
used for fixed prosthodontics proce-
dures.1 Some clinicians prefer sideless
plastic trays, while others use metal
trays with side walls.2 Proponents for
the plastic trays claim they are easy to
use, while others believe they are too
flexible. The use of a sideless plastic
tray provides an escape of impression
material during impression making.1
The mesh designs of double-arch
trays differ.3 Some are filamentous,
while others are crosshatched. It ap-
pears that all meshes are made from
a relatively impermeable nylon type
of material that does not absorb flu-
ids. A filamentous mesh is thin and
adapts readily to occlusal surfaces.
Crosshatched meshes can resist full
closure because the mesh does not
adapt as readily to the occlusal sur-
faces. The meshes are compressible to
provide the closest possible contact,
to afford accurate interocclusal regis-
tration.
The impression material of choice
when using dual-arch trays is vinyl
polysiloxane, because it is clean, pro-
vides excellent surface detail, is ac-
curate, and is easy to remove.4 Vinyl
polysiloxane is a dual-mix rigid mate-
rial that sets hard; the flexible tray be-
comes semirigid upon polymerization
of the impression material. There may
be some resistance to closure when
using various impression trays of this
type because they contain different
impression support systems. The pur-
pose of this study was to compare 3
tray systems and to determine wheth-
er closure into double-arch trays was
affected by impression volume and/
or type of tray support, for example,
webbing or mesh. The null hypothesis
was that there would be no significant
difference found between tray design
and registration of maximum inter-
cuspation.
MATERIAL AND METHODS
Two impact-resistant casts were
mounted in maximum intercuspal po-
sition on a semi-adjustable articulator
(Artex AL; Amann Girrbach AG, Ko-
Hahn et al
blach, Austria). The casts were made
of epoxy resin die material, which is
more resistant to abrasion and stron-
ger than a gypsum die material. The
casts were developed by P.K. Thomas
in 1960. The methodology of using
these casts was established by Hatzi
et al in 2001.5 Impressions of the left
posterior quadrant of the casts were
made using plastic double-arch trays.
The 2 independent variables were
type of tray (from 3 different manu-
facturers) and volume of impression
material. A sample size of 10 was
used, yielding 60 impressions.
Three common types of dual-
arch trays were used, and all were
plastic and comparable in size: First
Bite, with thin, gauze, nylon web-
bing (Dentsply Caulk, Milford, Del),
3-Way, with a nylon double cross-
hatch mesh (Sultan Dental Materials
Ltd, Englewood, NJ), and Premium
3-in-1 Tray, made of a single cross-
hatch material (Premium Plus Dental
Supplies, Inc, Brooklyn, NY) (Fig. 1).
Using a vernier caliper (General Tools
& Instruments Co, New York, NY), the
thickness of Sultan’s 3-Way mesh was
measured to be 0.25 mm, and both
First Bite and 3-in-1 trays measured
0.10 mm.
The dual-arch trays were placed
to ensure that the casts mounted on
the articulator could be closed into
occlusion without any interference
from the tray.6 Vinyl polysiloxane im-
pression material (Aquasil Ultra Rigid
Fast Set; Dentsply Caulk) was auto-
1 Comparison of 3 dual-arch impression trays: Dentsply Caulk First
Bite (left), Sultan 3-Way (middle), and Premium 3-in-1 Tray (right).
363
mixed and injected into a dual-arch
tray using a disposable syringe tip.
A dispensing gun (Dispensing Gun;
Dentsply Caulk) was used to produce
a constant volume with each pull
of the trigger; the volume is not de-
pendent on the force applied to the
trigger. For the low volume group, 5
complete activations of the automix
cartridge were used to dispense tray
material into both sides of the dual-
arch tray. The first complete activa-
tion was not enough to bring the
material beyond the disposable tip.
Then, 2 more complete activations
were dispensed on the maxillary side,
and the remaining 2 complete activa-
tions were dispensed on the mandib-
ular side. For the high volume group,
7 complete activations were used, 1
preliminary pull, 3 on the maxillary
side, and 3 on the mandibular side.
After each record was made, it was
weighed with a balance scale (Vicon
VIC-711 Precision Balance scale; Ac-
culab Sartorius Group, Edgewood,
NY) with 0.1-g readability. The mass
of the appropriate dual-arch tray was
subtracted, and the volume was cal-
culated by dividing the mass by the
density of the vinyl polysiloxane im-
pression material (Aquasil Ultra Rigid
Fast Set; Dentsply Caulk), which is
1.5 g/ml, as purported by the manu-
facturer. Five complete activations
equaled 5.4 ±0.1 ml, and 7 complete
activations equaled 8.3 ±0.1 ml. The
dual-arch trays were placed onto the
mandibular left quadrant, the articu-
364
2 Impression tray held in fixed position with articulated typodont in closed
position (1.2-kg weight is not shown). Incisal pin was maintained in place.
3 Example of zero (left), low (middle), and high (right) volume.
lator was closed, and a 1.2-kg weight
was placed on top to simulate a con-
stant occlusal force.7 The weight of
1.2 kg was chosen because of its avail-
ability and to consistently reproduce
the same amount of force during clo-
sure. A clamp/jig was used so that the
position of the impression tray was
constant and reproducible between
trials (Fig. 2). The impressions were
removed 5 minutes after the articula-
tor was closed (Fig. 3).
The method of recording and mea-
suring was as follows. A silicone re-
cord (Regisil Bite Registration Mate-
rial; Dentsply Caulk) of contact, near
contact, or noncontact was made.
The Journal of Prosthetic Dentistry
The record was a 3-dimensional (3-D)
representation of the specific cusp/
fossa relationship. A silicone calibra-
tion grid was also made at the time
of record making.8 The calibration
grid was formed by placing a small
amount of silicone on the lower plate
of a press (made by the second au-
thor), the upper plate of which con-
tained a small sphere with a diameter
of three eighths of an inch. The press
was closed until contact was made
between the sphere and plate. The
press provided a uniformly circular
silicone grid that was used for calibra-
tion purposes.8 The grid and record
were illuminated by placing them on
Volume 102 Issue 6
a direct current light box (Fostec;
Schott North America, Inc, South-
bridge, Mass) and photographed at
a fixed distance of 8 inches (Fig. 4).
Direct current was used because it
provides a constant emission of light.
If the records were placed on the light
box at varying angles, this would po-
tentially cause variation in light trans-
mission. It was estimated through
geometric formulas that if the angle
was less than 25 degrees, the uncer-
tainty would be less than 10% of the
range of interest.5 Excessive material
from the interocclusal records was re-
moved to minimize placement error.5
The gray scale digital camera
(CFW-1308M; Scion Corp, Frederick,
Md) was set at a fixed distance of 8
inches and captured the light trans-
mitted through the impression mate-
rial. The camera transferred this infor-
mation to an image analysis program
(ImageJ; NIH, Bethesda, Md).8 This
system allows the different amounts
of light projected through the im-
pression to be translated into a gray
scale value which can be assigned
a thickness value. In this study, the
values of the picture elements (pix-
els) were determined by the optical
density of the impression material.8
The effective thickness assigned to a
pixel corresponded to the thickness
at the center of the pixel.8 An image
was made of the calibration grid on
the light box in the same manner as
for the impressions. A thin rectangu-
lar box was selected across the diam-
eter of the circle; the ends of the rect-
angular box were dark gray, and the
center, from which point most of the
light was transmitted, was white. At
any point along the rectangle, a pixel
could be assigned a gray scale value
(GSV), which signified how black or
white that pixel was. Using sphere
geometry, the thickness of the cali-
bration grid was determined and was
correlated to its assigned GSV. This
allowed for the determination of the
corresponding thickness of any GSV.
One way to describe a computer-
ized record is to describe it numerical-
ly. The mesiolingual cusp on the max-
Hahn et al
December 2009
4 Calibration grid (left) and impression (right) photographed on light box.
800
700
600
Thickness (µm)
500
400
300
200
100
0
0 50
5 Relationship between gray scale value and corresponding thickness.
illary left first molar was highlighted
with a customized standard circular
template for all 60 records. Through
the image analysis software (ImageJ;
NIH), a customized 3-mm-diameter
circle was used for each specimen.
The same circle was applied to each
specimen so that the exact same num-
ber of pixels was analyzed. The circle
was centered over the mesiolingual
cusp using radiolucent landmarks
that were visible on every image cap-
tured. Then, those regions of interest
(ROIs) were analyzed for an average
of all pixels. The range was 15 (black)
to 255 (white). All of the ROIs were
represented by a number, which was
lower if less light was transmitted (a
record with more noncontact areas).
As described above, the use of the grid
established the relationship between
Hahn et al
100 150 200
Gray Scale Value
transmitted light (GSV) and thickness
of the impression material8 (Fig. 5).
To ensure reliability of the re-
cord-making technique, 10 control
specimens of impression material
dispensed directly onto the articu-
lated typodont were collected. Two
pulls were distributed onto 3 occlusal
surfaces from the mandibular second
premolar to the second molar. The
same customized circular template
captured the mesiolingual cusp of
the maxillary first molar for analy-
sis. To assess for variability between
groups, each type of tray was placed
on the light box without any impres-
sion material, and the light transmis-
sion was measured. A 2-way analysis
of variance (ANOVA) was used to test
main and interaction effects (tray (3)
by volume (2)), and Tukey Honestly
365
250
Significant Difference (HSD) post hoc
test was used to evaluate significant
differences among groups.
RESULTS
The control group’s regions of in-
terest (ROIs) were measured, and the
mean was 248.9 ±2.4 gray scale units.
This indicated that the contact point
was close to the maximum white gray
scale value, which was 255 GSV. The
purpose of the control group was
to assess reliability of the entire ap-
paratus, from impressions to image
analysis. For this reason, it was not in-
cluded in the statistical analysis with
the impressions with trays. When the
trays without impression material
were measured for light transmission,
all of them were 255 GSV. The photos
366
200
180
Low volume
160
Light (Gray Scale Units)
High volume
140
120
100
80
60
40
20
0
First Bite
Tray Type
6 Mean light in gray scale units transmitted for each tray at
2 different volumes.
were made at the same light box in-
tensity as the impressions with trays.
The results from the 3 (tray) by 2
(volume) between-subjects ANOVA
include a significant main effect for
the tray F (2, 54) = 323.6; P<.001.
No significant main effect was found
for volume F (1, 54) = 0.15; or tray
by volume interaction F (2, 54) =1.8.
The Tukey HSD test for equal sample
sizes was used to examine the signifi-
cant differences among tray types.
The Premium tray type had the high-
est mean (SD) GSV, 179.8 (8.1), sig-
nificantly higher than First Bite (164.7
(19.8); P<.001) or Sultan (82.8 (7.4);
P<.001) trays. The Sultan tray had
the lowest gray scale value, which was
significantly lower (P<.001) than First
Bite (Fig. 6). Figure 5 depicts the re-
lationship between gray scale value
and thickness of the contact point,
measured in micrometers. First Bite
was 250-300 μm, Sultan was 430-
530 μm, and Premium was 250-300
μm. The mean (SD) control measured
248.9 (2.4) gray scale units, which
translates to being less than 200 μm
thick in the same region of interest.
First Bite and Premium were at least
50 μm thicker, and Sultan was more
than 200 μm thicker than an impres-
sion with no tray.
The Journal of Prosthetic Dentistry
Sultan Premium
DISCUSSION
Interestingly, the thickness of vinyl
polysiloxane impression material was
significantly different among the 3
tray types. Therefore, the null hypoth-
esis was rejected. All of the trays were
sideless and comparable in size. The
only differences between trays were
the mesh design and mesh thickness.
The thicker meshes were more stiff
and resisted complete closure. Such
shifts in closure may affect accuracy
in the transfer to an articulator. The
region of interest analyzed was the
mesiolingual cusp of the first molar.
First Bite and Premium both mea-
sured 250-300 μm for this area. Sul-
tan was thicker, at 430-530 μm. Com-
plete closure did not occur with any
of the trays when compared to the
control. The aim of dual-arch trays
is to be as thin as possible to allow
intimate interocclusal contact. The
double crosshatched webbing signifi-
cantly hindered closure.
Increasing the volume by 3 ml did
not affect closure. The null hypothesis
stated that no difference was expect-
ed between the low volume group and
high volume group. Therefore, the
null hypothesis was not rejected with
regard to volume. Assuming that the
1.2-kg weight was applied at a con-
Volume 102 Issue 6
stant rate for each trial, closure would
be dependent on the viscosity of the
material. The next step would be to
examine the differences between vi-
nyl polysiloxane impression materials
with different viscosities to determine
if heavy-body materials hinder inter-
occlusal contact more than medium-
body materials. Although there was
no difference in vitro, there may be a
difference clinically, because patients
occlude with different amounts of
force. The mesh is compressible, and,
when coupled with the presence of
saliva and different closure forces, the
results may be different.
Different trays were compared in
this study because, to date, it is un-
known how trays differ once impres-
sion material and closure force are
added. This study could be improved
by using a micrometer screw to mea-
sure the exact thickness of the meshes
studied. In addition, thin articulating
paper should be used to ensure that
the incisal pin contacts the table af-
ter each impression is made. Further-
more, a wider range of weights should
be used to resemble forces that are
clinically applicable. Finally, to further
validate the results, stone casts made
from the impressions should be evalu-
ated for accuracy.
CONCLUSIONS
This in vitro study compared the
effect of impression material vol-
ume and tray design on closure of
mounted casts. Sultan’s thicker dou-
ble crosshatch mesh significantly hin-
dered closure, more than the tray de-
signs of First Bite or Premium trays.
Increasing the amount of impression
material did not hinder interocclusal
closure.
REFERENCES
1. Breeding LC, Dixon DL. Accuracy of casts
generated from dual-arch impressions. J
Prosthet Dent 2000;84:403-7.
2. Kaplowitz GJ. Trouble-shooting dual
arch impressions II. J Am Dent Assoc
1997;128:1277-81.
Hahn et al
December 2009
3. Wilson EG, Werrin SR. Double arch impres-
sions for simplified restorative dentistry. J
Prosthet Dent 1983;49:198-202.
4. Parker MH, Cameron SM, Hughbanks JC,
Reid DE. Comparison of occlusal con-
tacts in maximum intercuspation for two
impression techniques. J Prosthet Dent
1997;78:255-9.
5. Hatzi P, Millstein P, Maya A. Determining
the accuracy of articulator interchangeabil-
ity and hinge axis reproducibility. J Prosthet
Dent 2001;85:236-45.
367
6. Ceyhan J, Johnson G, Lepe X. The effect of Corresponding author:
tray selection, viscosity of impression mate- Sara Hahn
rial and sequence of pour on the accuracy Harvard School of Dental Medicine
of dies made from dual-arch impressions. J 188 Longwood Ave
Prosthet Dent 2003;90:143-9. Boston, MA 02115
7. Davis RD, Schwartz RS. Dual-arch and cus- Fax: 617-432-0901
tom tray impression accuracy. Am J Dent E-mail: sara_hahn@hsdm.harvard.edu
1991;4:89-92.
8. Millstein PL. A method to determine occlu- Acknowledgements
sal contact and noncontact areas: prelimi- The authors thank Dentsply Caulk for provid-
nary report. J Prosthet Dent 1984;52:106- ing impression materials and Deborah Lukey
10. for her technical assistance.

Copyright © 2009 by the Editorial Council for

The Journal of Prosthetic Dentistry.

Noteworthy Abstracts of the Current Literature

Changes in patient and FDP profiles following the introduction of osseointegrated im-
plant dentistry in a prosthodontic practice

Walton TR.
Int J Prosthodont 2009;22:127–35.

Purpose: This paper describes the effects of implant-related dentistry on patient profiles and the types of fixed dental
prostheses provided in clinical practice.

Materials and Methods: All implant- and tooth-supported prostheses provided in a prosthodontic practice between
1984 and 2007 were tabulated. Incidence was analyzed in relation to gender, age at time of prostheses insertion, and
prostheses type.

Results: Tooth-supported single crowns (TSCs) and tooth-supported fixed dental prostheses (TFDPs) were involved
in 97% of treatments requiring fixed dental prostheses from 1984 to 1991. From 1991 to 2007, however, a marked
increase in the number of implants restored directly corresponded with a decrease in the number of TFDPs so that by
2007, implant-supported fixed dental prostheses (IFDPs) accounted for 81% of all tooth replacements. Between 1984
and 2007 the incidence of TFDPs was 61% in females and 39% in males, whereas the incidence of IFDPs was 55% in fe-
males and 45% in males. IFDPs were also involved in 35% of restorations in patients under 31 years of age and TFDPs
in 19%. In the under-21 years age group, IFDPs were more common in females (9%) than males (4%), but in the 21 to
30 years age group they were more common in males (21%) than females (13%). There was a decrease in three-unit
TFDPs, in TFDPs with four or more pontics and those not satisfying Ante’s Law, and in teeth used that had been sub-
jectively assessed to have an unfavorable 10-year prognosis at the time of prosthesis insertion.

Conclusions: The incorporation of osseointegrated implant dentistry into a clinical practice has resulted in changes in
the patient profile and type of fixed dental prostheses provided, including a decrease in the use of TFDPs; an increase
in the referral of patients under 31 years of age; a decrease in three-unit, long-span, and complex TFDPs; and a de-
crease in tooth abutments assessed to be structurally or biologically compromised.

Reprinted with permission of Quintessence Publishing.

Hahn et al