Marquette University

e-Publications@Marquette
School of Dentistry Faculty Research and
Dentistry, School of
Publications

1-1-1996

Differential Scanning Calorimetry (DSC) Analyses

Of Superelastic And Nonsuperelastic Nickel-
Titanium Orthodontic Wires
T. Gerard Bradley
Marquette University, [email protected]

William A. Brantley
Ohio State University - College of Dentistry

Bill M. Culbertson
Ohio State University - College of Dentistry

American Journal of Orthodontics and Dentofacial Orthopedics, Vol 109, No. 6 (1996): pg. 589-597.
DOI. © 1997 Elsevier. Used with permission.
Dr. T. Gerard Bradley was associated with Ohio State University - College of Dentistry at time of
publication.
NOT THE PUBLISHED VERSION; this is the author’s final, peer-reviewed manuscript. The published version may be
accessed by following the link in the citation at the bottom of the page.

Differential Scanning Calorimetry

(DSC) Analyses Of Superelastic And
Nonsuperelastic Nickel-Titanium
Orthodontic Wires

Thomas Gerard Bradley

College of Dentistry, The Ohio State University
Columbus, OH
William A. Brantley
College of Dentistry, The Ohio State University
Columbus, OH
Bill M. Culbertson
College of Dentistry, The Ohio State University
Columbus, OH

Abstract: The purpose of this study was to determine the transformation

temperatures for the austenitic, martensitic, and rhombohedral (R) structure
phases in representative as-received commercial nitinol (NiTi) orthodontic
wire alloys, to reconcile discrepancies among recent publications. Specimens
were examined by differential scanning calorimetry (DSC) over a temperature
range from approximately −170° C to 100° C, with a scanning rate of 10° C
American Journal of Orthodontics and Dentofacial Orthopedics, Vol 109, No. 6 (1996): pg. 589-597. DOI. This article is ©
Elsevier and permission has been granted for this version to appear in e-Publications@Marquette. Elsevier does not grant
permission for this article to be further copied/distributed or hosted elsewhere without the express permission from
Elsevier.
1
NOT THE PUBLISHED VERSION; this is the author’s final, peer-reviewed manuscript. The published version may be
accessed by following the link in the citation at the bottom of the page.

per minute. Two different pathways, with the intermediate R structure either
absent or present, were observed for the transformation from martensitic to
austenitic NiTi, whereas the reverse transformation from austenitic to
martensitic NiTi always included the R structure. The enthalpy (ΔH) for the
transformation from martensite to austenite ranged from 0.3 to 35 calories
per gram. The lowest ΔH value for the nonsuperelastic Nitinol wire is
consistent with a largely work-hardened, stable, martensitic microstructure in
this product. The DSC results indicate that the transformation processes are
broadly similar in superelastic, body-temperature shape-memory, and
nonsuperelastic NiTi wires. Differences in bending properties for the NiTi
orthodontic wires at room temperature and 37° C are due to the relative
proportions of the metallurgical phases in the microstructures.

Commercial nickel-titanium orthodontic wires fabricated from

near-equiatomic NiTi (nitinol) compositions were introduced during the
1970s, after the pioneering research of Andreasen and coworkers.1,2
The mechanical properties and clinical characteristics of the original
NiTi product, marketed as Nitinol by the Unitek Corporation, were
reported by Andreasen and Morrow.3 This wire has a very low elastic
modulus, as well as an extremely wide working range and excellent
spring-back (using a 12.5 mm cantilever bending test span), thus
providing much lighter and more constant forces over an extended
treatment period, compared with the use of conventional stainless
steel wires. In additional publications, Andreasen and his colleagues
described the force delivery and temperature dependence for nitinol
wires.4-6

Subsequently, Burstone et al.7 and Miura et al.8 introduced the

Chinese and Japanese NiTi wires, respectively, to the specialty. These
wires, now marketed as Ni-Ti (Ormco/Sybron, Glendora, Calif.) and
Sentinol or Sentalloy (GAC International, Central Islip, N.Y.), displayed
superelastic characteristics, where the deactivation plot contains an
extended region of largely constant bending moment or tensile
stress.7,8 By using a cantilever bending test span (6 mm) similar to
interbracket distances, Khier et al.9 found that commercial NiTi
orthodontic wires could be classified as superelastic or nonsuperelastic,
where the latter products had much lower springback. The superelastic
wires were reported to be austenitic NiTi and underwent a reversible
stress-induced transformation to martensitic NiTi during activation,
whereas the nonsuperelastic wires, such as Nitinol, had stable work-
hardened martensitic structures.8 Later, Leu et al.10 employed
differential scanning calorimetry (DSC) to show that the phase

American Journal of Orthodontics and Dentofacial Orthopedics, Vol 109, No. 6 (1996): pg. 589-597. DOI. This article is ©
Elsevier and permission has been granted for this version to appear in e-Publications@Marquette. Elsevier does not grant
permission for this article to be further copied/distributed or hosted elsewhere without the express permission from
Elsevier.
2
NOT THE PUBLISHED VERSION; this is the author’s final, peer-reviewed manuscript. The published version may be
accessed by following the link in the citation at the bottom of the page.

transformations in the superelastic NiTi wires were more complex and

involved an immediate rhombohedral (R) structure.11,12

However, recent research has suggested that these concepts for

the metallurgical structures and phase transformations in orthodontic
NiTi alloys may require some revision. By using x-ray diffraction (XRD)
to analyze wires in the as-received condition and after tensile strain,
Thayer et al.13 identified XRD peaks for the martensitic NiTi structure
and concluded that the nonsuperelastic wires principally have the
austenitic structure at room temperature. This conclusion was
supported by the very low martensitic transformation temperatures
obtained by Chen et al.14 from electrical resistivity measurements,
where the Ms and Mf temperatures for the start and finish of this
transformation were typically below − 40° and − 100° C, respectively.
These transformation temperatures were much lower than those
obtained from DSC measurements for orthodontic alloys10,15,16 and
found in the materials science literature.17,18

The purpose of this study was to reconcile these conflicting

results by carefully examining the DSC heating and cooling curves for
several representative commercial NiTi wires. The aim was to
characterize the phase transformation processes in these different
products, specifically seeking evidence of very low temperature
martensite transformations. Specimens evaluated with the present
DSC apparatus could be cooled to approximately − 170° C, which was
not possible in the previous extensive study by Leu et al.10

Materials and methods

The commercial NiTi alloys selected for study are summarized in Table

I.

All wires were contributed by the manufacturers in standard

arch blank forms. Two of these products are superelastic (Nitinol SE,

American Journal of Orthodontics and Dentofacial Orthopedics, Vol 109, No. 6 (1996): pg. 589-597. DOI. This article is ©
Elsevier and permission has been granted for this version to appear in e-Publications@Marquette. Elsevier does not grant
permission for this article to be further copied/distributed or hosted elsewhere without the express permission from
Elsevier.
3
NOT THE PUBLISHED VERSION; this is the author’s final, peer-reviewed manuscript. The published version may be
accessed by following the link in the citation at the bottom of the page.

Unitek, Monrovia, Calif., and Ni-Ti),7-9 and the Nitinol wire (Unitek,
Monrovia, Calif.) is nonsuperelastic.3,9 Fletcher et al.16 have shown that
the Neo Sentalloy wire has shape-memory (superelastic) behavior at
37° C, but not at room temperature (22° C). The Titanal LT product
(Lancer Orthodontics, Carlsbad, Calif.) is advertised as a body-
temperature, heat-activated nickel-titanium wire with shape-memory
capability, and was thus expected to have characteristics very similar
to those for Neo Sentalloy. The clinically popular 0.016-inch diameter
size selected for Nitinol SE, Ni-Ti, and Nitinol has been used in other
studies7-9 of these wires. For the Titanal LT alloy, 0.018-inch diameter
wires were used due to the unavailability of the 0.016-inch diameter
size. For Neo Sentalloy (available only in rectangular cross-sections),
the popular 0.016 × 0.022-inch size was selected.

The wire specimens were carefully cut with a water-cooled,

diamond-embedded saw (Vari/Cut VC-50, Leco Corp., St. Joseph,
Mich.). Considerable care was taken in cutting the samples so that
minimal heat and stress would be generated. Straight segments of 5
mm length were cut from each arch blank, and five segments of each
wire were weighed (~20 mg sample size) with an electronic balance
and placed in an aluminum pan. The 5 mm segment length was
selected because of the pan dimensions, and pilot experiments
indicated that the use of five segments yielded excellent thermal
contact within the sample and DSC spectra of high resolution. Shorter
wire segments were not used to minimize the amount of cutting
required for sample preparation. There is no consensus in previous
studies about the ideal DSC sample weight for the NiTi wire segments,
and the present size was intermediate between those used by Todoroki
and Tamura (10 mg)17 and by Yoneyama et al. (40 mg).15

For the differential scanning calorimetry experiments (TA 910,

TA Instruments, Wilmington, Del.), each sample of five wire segments
was initially cooled from room temperature to −170° C and then
heated to 100° C at a scanning rate of 10° C/min. Subsequently, the
sample was cooled again to approximately −170° C at the same
scanning rate, and the same heating and cooling cycles were repeated
a second time. Pilot experiments established that excellent
reproducibility was achieved between nominally identical five-segment
samples of the same NiTi alloy, so that the procedure of a single
sample and replicate DSC heating and cooling runs was adopted for
American Journal of Orthodontics and Dentofacial Orthopedics, Vol 109, No. 6 (1996): pg. 589-597. DOI. This article is ©
Elsevier and permission has been granted for this version to appear in e-Publications@Marquette. Elsevier does not grant
permission for this article to be further copied/distributed or hosted elsewhere without the express permission from
Elsevier.
4
NOT THE PUBLISHED VERSION; this is the author’s final, peer-reviewed manuscript. The published version may be
accessed by following the link in the citation at the bottom of the page.

the present experiments. Dry nitrogen gas was used to flush the
specimen chamber, to prevent condensation of water vapor and
oxidation of the NiTi wires. A second, empty aluminum pan served as
an inert reference, and the apparatus was calibrated with an indium
standard, as well as by the melting point of water.

In the DSC procedure, the differential heat flow required to heat

or cool the experimental and reference samples at the same scanning
rate is recorded as a function of temperature to yield the spectrum or
thermogram. The start and finish temperatures of each phase
transformation were determined from tangent lines where the DSC
curve deviates from the adjacent baselines. Individual or combined
peak areas (enthalpies) were calculated from a peak on the DSC plot
whose area had previously been analyzed by the computer software.
Segments of beta-titanium arch wire and a polycrystalline alumina
bracket, which do not undergo phase transformations over this
temperature range, served as inert controls to establish the resolution
of the apparatus.

Results

Representative DSC curves for the NiTi wires studied are shown in Figs. 1 to

5.
Fig. 1. DSC heating and cooling curves for Nitinol SE (Unitek). H1 and H2
peaks are indicated for heating curve, and C1 and C2 peaks are shown for
cooling curve, along with possible position of very poorly resolved C3 peak.

American Journal of Orthodontics and Dentofacial Orthopedics, Vol 109, No. 6 (1996): pg. 589-597. DOI. This article is ©
Elsevier and permission has been granted for this version to appear in e-Publications@Marquette. Elsevier does not grant
permission for this article to be further copied/distributed or hosted elsewhere without the express permission from
Elsevier.
5
NOT THE PUBLISHED VERSION; this is the author’s final, peer-reviewed manuscript. The published version may be
accessed by following the link in the citation at the bottom of the page.

Apparent substructure in H2 peak is attributed to instrumentation noise. Area

under peak (enthalpy change ΔH for phase transformation) is determined
with respect to extrapolated baselines from adjacent portions of curve.

Fig. 2. DSC heating and cooling curves for Ni-Ti (Ormco). Curves are similar
in general appearance to those in Fig. 1Fig. 1Fig. 1, although phase
transformations occur over different temperature ranges. Fluctuations in
cooling curve below C3 peak are due to instrumentation noise.

Fig. 3. DSC heating and cooling curves for Neo Sentalloy (GAC). Note that
only one endothermic peak (H) is found on heating curve, in contrast to Figs.
1 and 2Figs. 1 and 2Figs. 1 and 2.

American Journal of Orthodontics and Dentofacial Orthopedics, Vol 109, No. 6 (1996): pg. 589-597. DOI. This article is ©
Elsevier and permission has been granted for this version to appear in e-Publications@Marquette. Elsevier does not grant
permission for this article to be further copied/distributed or hosted elsewhere without the express permission from
Elsevier.
6
NOT THE PUBLISHED VERSION; this is the author’s final, peer-reviewed manuscript. The published version may be
accessed by following the link in the citation at the bottom of the page.

Fig. 4. DSC heating and cooling curves for Titanal LT (Lancer). Small
fluctuations in curves are attributed to instrumentation noise. Phase
transformation characteristics are similar for those for Neo Sentalloy in Fig.
3Fig. 3Fig. 3.

Fig. 5. DSC heating and cooling curves for Nitinol (Unitek). Area under H peak
is much smaller than areas under peaks on heating curves for other NiTi
alloys (Table II).

American Journal of Orthodontics and Dentofacial Orthopedics, Vol 109, No. 6 (1996): pg. 589-597. DOI. This article is ©
Elsevier and permission has been granted for this version to appear in e-Publications@Marquette. Elsevier does not grant
permission for this article to be further copied/distributed or hosted elsewhere without the express permission from
Elsevier.
7
NOT THE PUBLISHED VERSION; this is the author’s final, peer-reviewed manuscript. The published version may be
accessed by following the link in the citation at the bottom of the page.

Two peaks (Figs. 1 and 2) or one peak (Figs. 3 to 5) were

observed on the heating curves, and these endothermic peaks have
been labeled as H1 and H2, or as H, respectively. From the analysis of
Todoroki and Tamura,17 the H1 and H2 peaks are interpreted as the
transformations from martensite to R structure (M-R) and from R
structure to austenite (R-A), while the H peak corresponds to the
direct transformation from martensite to austenite (M-A). Two
exothermic peaks (C1 and C2) observed on all cooling curves are
interpreted17 as the reverse transformations of austenite to R structure
(A-R) and R structure to martensite (R-M), respectively.

An additional low temperature peak (C3) appeared to occur near

the limit of resolution for Ni-Ti (Fig. 2) and Titanal LT (Fig. 4).
Locations of possible C3 for Nitinol SE (Fig. 1), Neo Sentalloy (Fig. 3),
and Nitinol (Fig. 5) have been placed near the cooling curves, but
these peaks cannot be resolved by DSC. If it exists, the C3 peak is
assumed to represent an additional phase transformation within the
complex martensitic structure.17

Table II summarizes the temperature ranges for all cooling

transformations, along with the temperatures and enthalpy changes
for the overall heating transformation from martensite to austenite.
Under the present experimental conditions, the temperatures in Table
II are considered to have an accuracy of approximately ± 2° C and the
enthalpy values have been rounded to the nearest 0.5 calories per
gram except for Nitinol. For Nitinol SE and Ni-Ti, the As, Af and ΔH
values correspond to both the H1 and H2 peaks in Figs. 1 and 2.

American Journal of Orthodontics and Dentofacial Orthopedics, Vol 109, No. 6 (1996): pg. 589-597. DOI. This article is ©
Elsevier and permission has been granted for this version to appear in e-Publications@Marquette. Elsevier does not grant
permission for this article to be further copied/distributed or hosted elsewhere without the express permission from
Elsevier.
8
NOT THE PUBLISHED VERSION; this is the author’s final, peer-reviewed manuscript. The published version may be
accessed by following the link in the citation at the bottom of the page.

Highly approximate Ms2 temperatures have been estimated for the C3

peaks, and no Mf2 temperature could be resolved for any alloy.

Referring to Figs. 1 and 2, the two peaks in both the heating

and cooling curves for Nitinol SE and Ni-Ti indicate that the R structure
forms as an intermediate phase during the reversible M-A
transformation. The phase transformation temperatures differed
somewhat for these two alloys, as shown in Table II. Thermal analyses
of the two controls (beta-titanium and alumina) indicated that the
substructure in the H2 peak for Nitinol SE and the fluctuations on the
cooling curve below the C3 peak for Ni-Ti were due to instrumentation
noise.

The curves for Neo Sentalloy in Fig. 3 show transformation

temperature ranges similar to those reported by Fletcher et al.,16
except that an unresolved pair of peaks for the M-R and R-A
transformations were observed in the earlier study. Although
exhibiting some instrumentation noise, the curves for Titanal LT in Fig.
4 resemble those for Neo Sentalloy, and the transformation
temperatures were similar. During cooling, the R structure has a role
in the A-M transformation for these shape-memory alloys, but
apparently not for the heating cycle where only a single H peak was
observed. Lastly, for the nonsuperelastic wire Nitinol a single broad,
weak peak occurred during the heating cycle, and two weak peaks
were observed during the cooling cycle.

Discussion

When the present DSC phase transformation temperatures in

Table II are compared with those determined by Chen et al.14 from the
temperature dependence of electrical resistivity, there is general
agreement for the two NiTi wires (Ormco Ni-Ti and Nitinol) common to
both studies. The present As, Af and Rs temperatures for Nitinol are
close to the temperatures reported by Chen et al. (who did not provide
Rf temperatures). For the Ni-Ti alloy, Chen et al. reported differences
in the phase transformation temperatures for the original Chinese NiTi7
and the Ormco Ni-Ti product,9 which may be attributed to differences
in the two manufacturing processes. Fariabi et al.18 have shown that
variations in the amount of cold work and the heat treatment
temperature can have substantial effects on the phase transformation
American Journal of Orthodontics and Dentofacial Orthopedics, Vol 109, No. 6 (1996): pg. 589-597. DOI. This article is ©
Elsevier and permission has been granted for this version to appear in e-Publications@Marquette. Elsevier does not grant
permission for this article to be further copied/distributed or hosted elsewhere without the express permission from
Elsevier.
9
NOT THE PUBLISHED VERSION; this is the author’s final, peer-reviewed manuscript. The published version may be
accessed by following the link in the citation at the bottom of the page.

temperatures for a near-equiatomic NiTi alloy. The present data in

Table II generally show better agreement with the Chen et al.
transformation temperatures for Chinese NiTi. Comparison of the
present results for Titanal LT is not possible, since Chen et al. used the
original Titanal9 product in their study.

Excellent agreement was found for the as-received Ni-Ti alloy

between the present C1 peak in Fig. 2 and the single peak observed by
Leu et al.10 during cooling. This latter study used an apparatus (DSC-3
Perkin-Elmer, Norwalk, Conn.) where specimens were cooled by dry
ice, and the minimum useful temperature on the DSC plots was
approximately 0° C. There was also excellent agreement between the
H2 peak in Fig. 2 and the single peak on their heating DSC curve. Any
lower temperature peak on their heating curve would have been
inaccessible and at substantial variance with the H1 peak in Fig. 2. The
present C2 peak was also inaccessible with their apparatus.

Although the usual phase transformations in cold-worked near-

equiatomic nonorthodontic NiTi alloys involve one or two peaks on the
heating and cooling DSC curves,17,18 Todoroki and Tamura17 also
observed three peaks on the cooling curves under certain heat
treatment and thermal cycling conditions. They noted that other
investigators had proposed the occurrence of a complex series of
martensitic reactions in NiTi alloys. Todoroki and Tamura also found
that electrical resistivity measurements provided much lower
resolution than DSC for the study of phase transformations in NiTi
alloys and that the two techniques did not yield the same
transformation temperature ranges and peaks. An additional
advantage of DSC is that the enthalpy change for a phase
transformation can be determined, which is not possible with electrical
resistivity measurements.

The present observation that the nonsuperelastic Nitinol wire is

largely martensitic at room temperature (Fig. 5) is in agreement with
other investigators,8,9,15 including x-ray diffraction studies by Khier et
al.19 However, this conclusion is contradictory to Thayer et al.,13 whose
XRD experiments with wires subjected to tensile strain suggested that
as-received Nitinol has the austenitic structure at room temperature.
Certain martensite variants (orientations)20 should be favored when
NiTi orthodontic wires undergo substantial tensile strain. Thus the
American Journal of Orthodontics and Dentofacial Orthopedics, Vol 109, No. 6 (1996): pg. 589-597. DOI. This article is ©
Elsevier and permission has been granted for this version to appear in e-Publications@Marquette. Elsevier does not grant
permission for this article to be further copied/distributed or hosted elsewhere without the express permission from
Elsevier.
10
NOT THE PUBLISHED VERSION; this is the author’s final, peer-reviewed manuscript. The published version may be
accessed by following the link in the citation at the bottom of the page.

martensite peaks observed by Thayer et al. after tensile loading may

be different from those in the as-received wires studied by Khier et al.,
where the work-hardened martensitic structure and preferred
crystallographic orientation resulting from the manufacturing
processes would yield different combinations of variants. Research is
needed to establish the metallurgical structures of orthodontic NiTi
wires and coil springs21 for the as-received condition and after heat
treatment (conveniently performed by the direct electrical resistance
technique16,22), and with clinical use where practical levels of
deformation are encountered.

Although phase transformation processes in NiTi orthodontic

wires are more conveniently studied by DSC than XRD, the DSC
results in this study have demonstrated the complexity of the
transformation processes between austenite and martensite in the
various types of commercial NiTi orthodontic wires. Otsuka12 has
discussed the R structure, which appears to be an intermediate phase
for both the M-A and A-M transformation directions in Nitinol SE and
Ni-Ti but seems to only participate in the A-M transformation path in
Neo Sentalloy, Titanal LT, and Nitinol. The broad H peak for Nitinol
may correspond to two successive transformations involving the R
structure, but there was insufficient DSC resolution of this very low
energy peak.

Although there is an energy difference between the overall M-A

and A-M transformations in NiTi alloys,20 the total enthalpy changes
associated with the peaks on the heating and cooling curves in Figs. 1
to 5 appear to be similar. However, such comparisons are approximate
because of the difficulty in accurately defining the DSC peak positions,
particularly any C3 peak. Lee et al.23 found that the enthalpy change
for the exothermic A-M transformation was about 60% of that for the
endothermic M-A transformation in NiTi alloys; these investigators
observed only a single peak in the cooling and heating DSC curves.

Hysteresis, the temperature difference between the forward and

reverse transformations in NiTi alloys,20 is evident when corresponding
peaks on the heating and cooling curves are compared. The C1 peak
(A-R) occurs at lower temperatures than the H2 peak (R-A) in Ni-Ti in
Fig. 2, or than the H peaks (M-A) for Neo Sentalloy and Titanal LT in
Figs. 3 and 4. However, there is little difference in the temperatures of
American Journal of Orthodontics and Dentofacial Orthopedics, Vol 109, No. 6 (1996): pg. 589-597. DOI. This article is ©
Elsevier and permission has been granted for this version to appear in e-Publications@Marquette. Elsevier does not grant
permission for this article to be further copied/distributed or hosted elsewhere without the express permission from
Elsevier.
11
NOT THE PUBLISHED VERSION; this is the author’s final, peer-reviewed manuscript. The published version may be
accessed by following the link in the citation at the bottom of the page.

the C1 and H2 peaks for Nitinol SE in Fig. 1, and the C1 and H peaks
for Nitinol in Fig. 5. For all five alloys, there is a considerable
difference between the temperature ranges for the C2 peak (R-M) and
either the H1 (M-R) or H peaks. Moreover, there is no counterpart on
the heating DSC curves for the C3 peak.

Examination of several DSC plots published by Todoroki and

Tamura17 and by Fariabi et al.18 reveals that the areas under the R
structure peaks for the forward (R-A) and reverse (A-R)
transformations involving austenite can be either greater or less than
the areas under the forward (M-R) and reverse transformations (R-M)
involving martensite. The present ΔH values of 0.3 to 3.5 calories per
gram in Table II for the endothermic transformation from martensite
to austenite (H peak or the sum of the H1 and H2 peaks) are in good
agreement with recent values reported by Yoneyama et al.15 from DSC
studies of seventeen commercial NiTi alloys. These latter authors
measured values of ΔH (typically for the combined M-R and R-A
transformations) ranging from 1.0 to 4.9 calories per gram, but
reported no endothermic M-A transformation for Nitinol similar to the
very weak peak in Fig. 5. Fariabi et al.18 found that the endothermic
ΔH values ranged from 3.6 to 5.7 calories per gram for cold-worked
and heat-treated nonorthodontic alloys.

Comparison of phase transformations for orthodontic NiTi wires

with NiTi alloys in the materials science literature is hindered by their
dependence on the amount of cold work, heat treatment temperature,
and Ni/Ti ratio. Todoroki and Tamura17 and Fariabi et al.18 have
presented extensive data on the complex influences of cold work and
heat treatment on the transformation pathways and enthalpy changes.
These two critical factors are expected to vary among the different
manufacturers for orthodontic NiTi products and would account for the
differences observed by Chen et al.14 for Chinese NiTi and the
equivalent Ormco Ni-Ti. Melton and Mercier24 have shown that varying
the alloy composition from 54.5% to 55.5 wt% nickel causes the Ms
temperature to decrease from 57° C to −30° C, with a decrease in the
hysteresis between the forward and reverse transformations from 55°
to 32°. Thus small variations in the manufacturing procedures for the
NiTi wires can have enormous effects on the phase transformations.
The incompletely resolved H1 and H2 peaks previously observed by
Fletcher et al.16 for Neo Sentalloy could be due to manufacturing
American Journal of Orthodontics and Dentofacial Orthopedics, Vol 109, No. 6 (1996): pg. 589-597. DOI. This article is ©
Elsevier and permission has been granted for this version to appear in e-Publications@Marquette. Elsevier does not grant
permission for this article to be further copied/distributed or hosted elsewhere without the express permission from
Elsevier.
12
NOT THE PUBLISHED VERSION; this is the author’s final, peer-reviewed manuscript. The published version may be
accessed by following the link in the citation at the bottom of the page.

differences for the two batches of arch wires, although differences in

specimen preparation cannot be excluded. Considerable care was
taken to minimize stress and heating during preparation of the present
DSC samples since previous research has shown that both stress and
heat treatment at temperatures of 500° C and 600° C profoundly
affect the bending properties and metallurgical structures of the
superelastic wires.8-10 However, little effect was observed after the
heat treatment of nonsuperelastic wires.9

The heating DSC curves in Figs. 1 to 5 can be used to predict

the metallurgical phases in the as-received NiTi wires at room
temperature and 37° C. For the two room temperature superelastic
wires, Nitinol SE and Ni-Ti, the conversion to austenite (H1 peak)
started at much lower temperatures than for the other three products
(H peak). For the two body-temperature shape-memory (superelastic)
wires, Neo Sentalloy and Titanal LT, the Af temperatures (H peak)
were below 37° C, although the As temperatures were much higher
than those for Nitinol SE and Ni-Ti. The heating DSC curves suggest
that Ni-Ti, Neo Sentalloy and Titanal LT should be almost or
completely austenitic in the oral environment, whereas Nitinol SE
should be a mixture of austenite and R structure. The nonsuperelastic
Nitinol should be entirely or almost entirely martensitic at room
temperature, although this product would be partially transformed to
austenite at mouth temperature. Alternatively, if the cooling DSC
curves are used to predict the phases, Ni-Ti, Neo Sentalloy and Titanal
LT would be austenitic at 37° C, whereas Nitinol SE and Nitinol would
be mainly R structure with some austenite present; at room
temperature, both Nitinol SE and Nitinol would be almost entirely R
structure. These predictions of phases from the DSC curves require
verification by some other technique, since the procedure of initially
cooling the samples to very low temperatures before commencing the
dynamic heating and cooling cycles may yield different proportions of
the metallurgical phases in these wires than those appropriate to
clinical conditions.

The foregoing DSC analyses of the metallurgical structures in

these representative alloys appear to suggest that the body-
temperature shape-memory NiTi wires are advantageous clinically in
comparison to the other superelastic and nonsuperelastic wires.
Although the nonsuperelastic wires have excellent springback and
American Journal of Orthodontics and Dentofacial Orthopedics, Vol 109, No. 6 (1996): pg. 589-597. DOI. This article is ©
Elsevier and permission has been granted for this version to appear in e-Publications@Marquette. Elsevier does not grant
permission for this article to be further copied/distributed or hosted elsewhere without the express permission from
Elsevier.
13
NOT THE PUBLISHED VERSION; this is the author’s final, peer-reviewed manuscript. The published version may be
accessed by following the link in the citation at the bottom of the page.

elastic ranges compared with orthodontic alloys other than nickel-

titanium,3,9,25 these NiTi wires have a work-hardened martensitic
structure and cannot take advantage of the transformation from
martensite to austenite that is required for superelastic and shape-
memory behavior. Andreasen and Hilleman1 originally stressed the
importance of the shape-memory property for clinical orthodontics,
noting that the temperature range for transition to austenite should lie
between room temperature and mouth temperature. Consequently, it
would appear that NiTi wires that are heat-sensitive in the oral
environment would provide a more clinically desirable combination of
light, continuous, and physiologic force levels. Use of refrigeration
would convert the shape-memory wires to the more deformable
martensitic structure, which may facilitate initial engagement into
severely malpositioned teeth under clinical conditions. Subsequent
thermal activation in the oral environment would transform these
wires to the fully austenitic structure, where light and continuous
forces are delivered with optimal springback characteristics compared
with the other types of commercially available products. Recently, a
new type of shape-memory wire has been introduced, where copper
has been incorporated in the NiTi alloy composition to alter the
transition temperature range (TTR) to austenite, lower or raising the
TTR between room and mouth temperature to modify the clinical
activity of the wire. If the TTR is above the temperature of the oral
environment, this wire would be active only when hot foods or liquids
are ingested and would thus be suitable for a treatment case where
limited activity is desired. Clinical trials comparing different types of
NiTi wires are required to verify these hypotheses.

Conclusions

Under the experimental conditions of this study, the following

conclusions can be drawn:

1. The five representative commercial NiTi orthodontic wires examined

undergo similar phase transformations over the temperature
range of −170° C to 100° C. A martensitic transformation peak
with a starting temperature from about −50° C to −80° C may
exist in these alloys, but was near the limit of resolution for the
DSC technique employed.

American Journal of Orthodontics and Dentofacial Orthopedics, Vol 109, No. 6 (1996): pg. 589-597. DOI. This article is ©
Elsevier and permission has been granted for this version to appear in e-Publications@Marquette. Elsevier does not grant
permission for this article to be further copied/distributed or hosted elsewhere without the express permission from
Elsevier.
14
NOT THE PUBLISHED VERSION; this is the author’s final, peer-reviewed manuscript. The published version may be
accessed by following the link in the citation at the bottom of the page.

2. The superelastic NiTi alloys (Nitinol SE and Ni-Ti) undergo austenitic

transformations involving the R structure which begin below 0°
C. The DSC analyses indicate that in the oral environment Ni-Ti
is almost entirely austenite, whereas Nitinol SE is a mixture of
austenite and R structure.
3. The nonsuperelastic alloy Nitinol is entirely or almost entirely
martensite at room temperature, and contains small additional
amounts of austenite in the oral environment.
4. The phase transformation characteristics for Neo Sentalloy and
Titanal LT are similar and consistent with their reported shape-
memory characteristics. These wires are entirely or almost
entirely austenite at body temperature.
5. The enthalpy changes measured by DSC for the transformations
from martensite to austenite were in good agreement with the
previously reported range for NiTi orthodontic wires. The
proportions of the metallurgical phases should be verified by
XRD or electron diffraction, or by the measurement of electrical
resistivity changes.

Acknowledgements

We thank Unitek/3M, Ormco/Sybron, GAC International, and Lancer

Orthodontics for contributing the arch wire products to this
investigation.

References

1
GF Andreasen, TB Hilleman. An evaluation of 55 cobalt substituted nitinol
wire for use in orthodontics. J Am Dent Assoc, 82 (1971), pp. 1373-
1375
2
GF Andreasen, PR Brady. A use hypothesis for 55 nitinol wire for
orthodontics. Angle Orthod, 42 (1972), pp. 172-177
3
GF Andreasen, RE Morrow. Laboratory and clinical analyses of nitinol wire.
Am J Orthod, 73 (1978), pp. 142-151
4
GF Andreasen, H Bigelow, JG Andrews. 55 nitinol wire: force developed as a
function of “elastic memory.” Aust Dent J, 24 (1979), pp. 146-149
5
G Andreasen. A clinical trial of alignment of teeth using a 0.019 inch thermal
nitinol wire with a transition temperature range between 31° C and
45° C. Am J Orthod, 78 (1980), pp. 528-537
6
G Andreasen, H Heilman, D Krell. Stiffness changes in thermodynamic nitinol
with increasing temperature. Angle Orthod, 55 (1985), pp. 120-126

American Journal of Orthodontics and Dentofacial Orthopedics, Vol 109, No. 6 (1996): pg. 589-597. DOI. This article is ©
Elsevier and permission has been granted for this version to appear in e-Publications@Marquette. Elsevier does not grant
permission for this article to be further copied/distributed or hosted elsewhere without the express permission from
Elsevier.
15
NOT THE PUBLISHED VERSION; this is the author’s final, peer-reviewed manuscript. The published version may be
accessed by following the link in the citation at the bottom of the page.

7
CJ Burstone, B Qin, JY Morton. Chinese NiTi wire—a new orthodontic alloy.
Am J Orthod, 87 (1985), pp. 445-452
8
F Miura, M Mogi, Y Ohura, H Hamanaka. The super-elastic property of the
Japanese NiTi alloy wire for use in orthodontics. Am J Orthod Dentofac
Orthop, 90 (1986), pp. 1M-10M
9
SE Khier, WA Brantley, RA Fournelle. Bending properties of superelastic and
nonsuperelastic nickel-titanium orthodontic wires. Am J Orthod
Dentofac Orthop, 99 (1991), pp. 310-318
10
L Leu, R Fournelle, W Brantley, T Ehlert. Evidence of R structure in
superelastic NiTi orthodontic wires. J Dent Res, 69 (1990), p. 313.
(IADR Abstracts)
11
D Goldstein, L Kabacoff, J Tydings. Stress effects on nitinol phase
transformations. J Metals, 39 (1987), pp. 19-26
12
K OtsukaIntroduction to the R-phase transition. TW Duerig, KN Melton, D
Stöckel, CM Wayman (Eds.), Engineering aspects of shape memory
alloys, : Butterworth-Heinemann Ltd, London (1990), pp. 36-45
13
TA Thayer, MD Bagby, RN Moore, R De Angelis. X-ray diffraction of NiTi
orthodontic archwires. J Dent Res, 72 (1993), p. 139. (IADR Abstracts)
14
R Chen, YF Zhi, MG Arvystas. Advanced Chinese NiTi alloy wire and clinical
observations. Angle Orthod, 62 (1992), pp. 59-66
15
T Yoneyama, H Doi, H Hamanaka, Y Okamoto, M Mogi, F Miura. Super-
elasticity and thermal behavior of Ni-Ti alloy orthodontic arch wires.
Dent Mat J, 11 (1992), pp. 1-10
16
ML Fletcher, S Miyake, WA Brantley, BM Culbertson. DSC and bending
studies of a new shape-memory orthodontic wire. J Dent Res, 71
(1992), p. 169. (AADR Abstracts)
17
T Todoroki, H Tamura. Effect of heat treatment after cold working on the
phase transformation in TiNi alloy. Trans Jpn Inst Metals, 28 (1987),
pp. 83-94
18
S Fariabi, PE Thoma, DN Abujudom. The effect of cold work and heat
treatment on the phase transformations of near equiatomic NiTi shape
memory alloy. Proc ICOMAT (1989)
19
S Khier, W Brantley, R Fournelle, T Ehlert. XRD and DSC studies of NiTi
orthodontic wire alloys. J Dent Res, 68 (1989), p. 386. (AADR
Abstracts)
20
CM Wayman, TW Duerig. An introduction to martensite and shape memory.
TW Duerig, KN Melton, D Stöckel, CM Wayman (Eds.), Engineering
aspects of shape memory alloys, : Butterworth-Heinemann Ltd,
London (1990), pp. 3-20
21
F Miura, M Mogi, Y Ohura, M Karibe. The super-elastic Japanese NiTi alloy
wire for use in orthodontics: part III, studies on the Japanese NiTi
alloy coil springs. Am J Orthod Dentofac Orthop, 94 (1988), pp. 89-96

American Journal of Orthodontics and Dentofacial Orthopedics, Vol 109, No. 6 (1996): pg. 589-597. DOI. This article is ©
Elsevier and permission has been granted for this version to appear in e-Publications@Marquette. Elsevier does not grant
permission for this article to be further copied/distributed or hosted elsewhere without the express permission from
Elsevier.
16
NOT THE PUBLISHED VERSION; this is the author’s final, peer-reviewed manuscript. The published version may be
accessed by following the link in the citation at the bottom of the page.

22
F Miura, M Mogi, Y Ohura. Japanese NiTi alloy wire: use of the direct electric
resistance heat treatment method. Eur J Orthod, 10 (1988), pp. 187-
191
23
JH Lee, JB Park, GF Andreasen, RS Lakes. Thermomechanical study of Ni-Ti
alloys. J Biomed Mater Res, 22 (1988), pp. 573-588
24
KN Melton, O Mercier. The mechanical properties of NiTi-based shape
memory alloys. Acta Met, 29 (1981), pp. 393-398
25
WA Brantley. Orthodontic wires. WJ O'Brien (Ed.), Dental materials:
properties and selection, : Quintessence Publishing, Chicago (1989).
Chap 20

American Journal of Orthodontics and Dentofacial Orthopedics, Vol 109, No. 6 (1996): pg. 589-597. DOI. This article is ©
Elsevier and permission has been granted for this version to appear in e-Publications@Marquette. Elsevier does not grant
permission for this article to be further copied/distributed or hosted elsewhere without the express permission from
Elsevier.
17