- Email: [email protected]
Thermal and mechanical characteristics of stainless steel, titanium-molybdenum, and nickel-titanium archwires
ORIGINAL ARTICLE
Thermal and mechanical characteristics of stainless steel, titanium-molybdenum, and nickel-titanium archwires Robert P. Kusya and John Q. Whitleyb Chapel Hill, NC Introduction: In recent years, nickel-titanium (Ni-Ti) archwires have been developed that undergo thermal transitions. Before the practitioner can fully utilize these products, the effect of those transitions within the clinical application must be understood. Methods: The transitional temperatures and mechanical stiffnesses of 3 archwire alloys—stainless steel, beta-titanium, and Ni-Ti—were investigated were for 7 products. Among the nickel-titanium alloys, 2 were thought to represent classic Ni-Ti products and 3 copper (Cu)-Ni-Ti products. By using 2 techniques, differential scanning calorimetry to measure heat flow and dynamic mechanical analysis to measure storage modulus, transition temperatures were evaluated from ⫺30°C to ⫹80°C. Results: With regard to the first technique, no transitions were observed for the stainless steel alloy, the beta-titanium alloy, and 1 of the 2 classic Ni-Ti products. For the other classic Ni-Ti product, however, a martensitic-austenitic transition was suggested on heating, and a reverse transformation was suggested on cooling. As expected, the Cu-Ni-Ti 27, 35, and 40 products manifested austenitic finish temperatures of 29.3°C, 31.4°C, and 37.3°C, respectively, as the enthalpy increased from 2.47 to 3.18 calories per gram. With regard to the second technique, the storage modulus at a low frequency of 0.1 Hz paralleled static mechanical tests for the stainless steel alloy (183 gigapascal [GPa]), the beta-titanium alloy (64 GPa), and the Nitinol Classic (3M Unitek, Monrovia, Calif) product that represented a stable martensitic phase (41 GPa). The remaining 4 Ni-Ti products generally varied from 20 to 35 GPa when the low-temperature or martensitic phase was present and from 60 to 70 GPa after the high-temperature or austenitic phase had formed. Conclusions: From the clinical viewpoint, the Orthonol (Rocky Mountain Orthodontics, Denver, Colo), Cu-Ni-Ti 27, Cu-Ni-Ti 35, and Cu-Ni-Ti 40 (SDS/Ormco, Glendora, Calif) products increased at least twofold in stiffness as temperature increased, best emulating the stiffness of Nitinol Classic below the transformational temperature and the stiffness of TMA (SDS/Ormco, Glendora, Calif) above the transformational temperature. Of the 3 Cu-Ni-Ti products, the least differences were found between Cu-Ni-Ti 27 and Cu-Ni-Ti 35, thereby questioning the justification for 3 similar products. (Am J Orthod Dentofacial Orthop 2007;131: 229-37)
P
revious investigators measured the tensile and flexure characteristics of archwires with differing outcomes, which depended on specific instrumentation, test methodology, and wire geometry.1-5 For example, stainless steel values varied from 22 ⫻ 106 to 30 ⫻ 106 psi,1,3 which in Systeme Internationale units corresponded to 150 to 200 gigapascal (GPa). Such differences profoundly influence the strength-torange ratio and consequently change the resilience by From the University of North Carolina, Chapel Hill. a Department of Orthodontics and Dental Research Center, School of Dentistry. b Dental Research Center, School of Dentistry. Supported in part by NIH grant 5 RO1 DE13201-03. Reprint requests to: Robert P. Kusy, University of North Carolina, DRC Building 210H, Room 313, CB#7455, Chapel Hill, NC 27599-7455; e-mail, [email protected]. Submitted, March 2005; revised and accepted, May 2005. 0889-5406/$32.00 Copyright © 2007 by the American Association of Orthodontists. doi:10.1016/j.ajodo.2005.05.054
the same proportions. Once either the working strength or its range is known along with the stiffness, the effectiveness of an archwire to move teeth is established.4,6,7 Primarily within the last 10 years, investigators have begun to characterize the transformational characteristics of archwires.8-12 Such studies are important for nickel-titanium (Ni-Ti) archwires whose reversible phase transformations occur from austenite to martensite, or sometimes an intermediate metastable R-phase occurs. Knowledge of the austenitic start and finish temperatures (on heating) and the martensitic start and finish temperatures (on cooling) establishes the stiffnesses of each archwire at ambient and oral-cavity conditions (34°C),13 and at the temperatures of ice cream (about 0°C) and hot coffee (about 60°C).14 Since the first commercial thermal analysis instrumentation in the 1960s, hardware and software capa229
230 Kusy and Whitley
bilities have grown so that today some 2 dozen techniques are available—2 of which are differential scanning calorimetry (DSC) and dynamic mechanical analysis (DMA). DSC measures heat flow as temperature changes.15 When changes occur from solid to liquid states (eg, the melting of ice) or, in the present case, from 1 phase to another in the same state (from martensite to austenite in the solid state), an endothermic (or exothermic) reaction occurs as a result of heating (or cooling). Thereby, the transitional temperatures and enthalpy changes associated with latent heat of transition can be measured.16 On the other hand, DMA measures the real component of the complex modulus—the storage modulus—as the temperature changes.17 When such a specimen is appropriately vibrated at 1 or more known frequencies, the response might not only lag behind but also become attenuated. These changes are most apparent as the temperature traverses a transformation such as a solid-state phase change that is found in some Ni-Ti products.18 In this article, the DSC and DMA techniques are used to determine both the phase transformations and the magnitude of storage moduli (a measure of wire stiffness) for 3 alloy groups. From these measurements, we show not only the moduli of invariant alloys but also the temperatures, enthalpies, and moduli of alloys undergoing solid-state phase transformations. Specifically, Ni-Ti products can exhibit storage moduli that vary with temperature by a factor of 2, when only small changes in composition and temperature occur. Therefore, Ni-Ti products should be selected on the basis of their properties and clinical applications—not on the basis of their costs and vendors. MATERIAL AND METHODS
Seven archwire products comprising 3 alloy groups were studied (Table I): a traditional austenitic stainless steel, the first-marketed titanium-molybdenum binary called beta-titanium, and 5 Ni-Ti alloys. Of the 5 Ni-Ti alloys, 2 were thought to be stabilized martensitic alloys in which processing prevented further transformations, and 3 were thought to be active alloys in which the amounts of copper (Cu) and chromium19,20 influenced the transformations from austenite to martensite on cooling (A ¡ M) and martensite to austenite on heating (M ¡ A).21,22 When both transformations were possible, the alloy was said to be reversibly thermoelastic, and the notation appears either as M ↔ A or, equivalently, as A ↔ M. In standard DSC mode, specimens (10-11 mg) of wire products were scanned 3 times with a differential scanning calorimeter (Q100, TA Instruments, New Castle, Del) and an attached refrigerated cooling sys-
American Journal of Orthodontics and Dentofacial Orthopedics February 2007
Table I.
Materials evaluated
Product
Alloy
Standard Rectangular TMA Nitinol Classic Orthonol
Stainless steel
Cu-Ni-Ti 27 Cu-Ni-Ti 35 Cu-Ni-Ti 40
Nickel-titanium Nickel-titanium Nickel-titanium
Beta-titanium Nickel-titanium Nickel-titanium
Constituents
Manufacturer
Fe, Cr, Ni, C
American Orthodontics Ti, Mo, Zr, Sn SDS/Ormco Ni, Ti 3M Unitek Ni, Ti Rocky Mountain Orthodontics Ni, Ti, Cu, Cr SDS/Ormco Ni, Ti, Cu, Cr SDS/Ormco Ni, Ti, Cu, Cr SDS/Ormco
All wire specimens were nominally 0.016 ⫻ 0.022 in. Mo, Molybdenum; Zr, zirconium; Sn, tin; Cr, chromium; Fe, iron; C, carbon; Ni, nickel; Cu, copper; Sn, tin; Ti, titanium.
tem (Fig 1, A). The instrument was calibrated with matched sapphire disks and then was run with indium. All specimens were scanned from ⫺30°C to 60°C under an atmosphere of ultrahigh purity nitrogen. Immediately after this temperature ramp, the specimens were scanned from 60 to ⫺30°C at the same rate. When both the specimen and the reference pans were heated at the programmable rate of 5C° per minute in the temperature-controlled chamber (Fig 1, B), changes in heat flow occurred when a phase transition happened. (In this article, there is an important difference between 5°C and 5C°. The former refers to a specific temperature on the Celsius scale; the latter indicates the magnitude of a thermal change that is independent of the actual Celsius temperature.) When the phase transition is a solid-solid transformation, an endothermic absorption peak is observed on heating that begins at the austenite start (AS) temperature and ends at the austenite finish (AF) temperature.23 This peak appears like other first-order transformations— eg, melting amd boiling peaks. After cooling at the same programmable rate, the transformation occurs somewhat later, and the phase change appears supercooled. That is, the transformation of A ¡ M occurs at the martensite start (MS) temperature, and the completion occurs at the martensite finish (MF) temperature (Fig 1, C ). After we measured these 4 temperatures, we calculated the endothermic peaks on heating and the exothermic peaks on cooling. These peaks are proportional to the enthalpies of heating (⌬HH) and cooling (⌬HC), respectively, and were calculated by using the sigmoidal tangent algorithm of Universal Analysis 2002 (TA Instruments). By using a 5-mm span, wire specimens were scanned in the 3-point bending fixture of a dynamic mechanical analyzer (2980, TA Instruments) outfitted with a liquid nitrogen gas cooling accessory (Fig 2, A). After preloading with 5 g at an amplitude of 0.0005 cm,
American Journal of Orthodontics and Dentofacial Orthopedics Volume 131, Number 2
Kusy and Whitley 231
Fig 1. DSC: A, Loading platform with centrally located reference pan and specimen pan; B, schematic of temperature-controlled chamber in vertical plane; C, typical heating and cooling thermograms defining 6 parameters of importance.
Fig 2. DMA: A, Three-point bending fixture; B, schematic of temperature-controlled chamber and its hardware in vertical plane; C, typical heating and cooling thermograms defining 7 parameters of importance.
the specimens were scanned by using frictionless air bearings and a drive motor from ⫺20° to 80°C at a rate of 1C° per minute. This was followed by a scan from 80° to ⫺30°C at the same rate in a temperature-
controlled chamber (Fig 2, B). The responses from 3 frequencies (0.1, 1.0, and 10 Hz) were sequentially measured at each temperature with an optical encoder. The lowest frequency best mimicked the static test
232 Kusy and Whitley
conditions of conventional tensile or bending tests. By using the same phase notation as the DSC tester, each storage modulus (which is directly proportional to the dynamic force applied) showed a similar relative delay of the cooling curve compared with its heating curve (Fig 2, C). This shift resulted in a hysteresis loop,22-25 which (as in hard vs soft magnets26) indicates the ease of the transformation. The low temperature phase (M) at temperature (T) ⬍MF has a lower storage modulus than the high temperature phase (A) at T ⬎AF. Consequently, on heating, the dynamic force is at its minimum for the martensitic phase as AS is approached and at its maximum for the austenitic phase once AF has been exceeded. On subsequent cooling, the A phase persists down to MS, but transformation to all M phase is not complete until MF. The horizontal magnitude associated with each hysteresis loop (⌬T) was calculated by first determining the storage modulus at the midpoint between AS and AF of the heating curve. Then, by extrapolating this modulus onto the cooling curve, the difference between the temperature of the heating and cooling traces, ⌬T, was obtained.
American Journal of Orthodontics and Dentofacial Orthopedics February 2007
Fig 3. Comparison of DSC thermograms of 4 archwires for which no transitions were expected: (top to bottom) Standard Rectangular, TMA, Nitinol Classic, and Orthonol (see Table I).
Statistical analysis
For both the DSC and DMA techniques, the means ⫾ 1 SD were determined for transition temperatures of all products that had first-order transformations. In both techniques, specific values were averaged: for DSC, representing the 3 runs; for DMA, these were the 3 frequencies. RESULTS
Figure 3 shows the DSC results, from top to bottom, of the heating and cooling thermograms for stainless steel, beta-titanium, and 2 Ni-Ti alloys, respectively. As expected, the Standard Rectangular (Am Orthodontics, Sheboygan, Wisc), TMA (SDS/Ormco, Glendora, Calif), and the Nitinol Classic (3M Unitek, Monrovia, Calif) products showed no transitions in the temperature regime of the oral cavity. Contrary to earlier beliefs and possibly because of a change in this Ni-Ti alloy, the Orthonol (Rocky Mountain Orthodontics, Denver, Colo) product had a small endothermic (or exothermic) peak on heating (or cooling).8 Whether this represents an M ¡ A transformation on heating and an A ¡ M transformation on cooling will be verified by the DMA results. Figure 4 details the 3 Cu-Ni-Ti products, which the manufacturer designates by their transitions of 27°C, 35°C, and 40°C (SDS/Ormco). When heated, all CuNi-Ti products had endothermic reactions in the order that their thermal monikers would imply. Two distinct transformations occurred for Cu-Ni-Ti 40, the interme-
Fig 4. Comparison of DSC thermograms in which at least 1 transition occurs on heating and cooling: (top to bottom) Cu-Ni-Ti 27, Cu-Ni-Ti 35, and Cu-Ni-Ti 40 (see Table I).
diate R-phase27 followed by the high-temperature or austenitic phase. When cooled, all had the appearance shown in Figure 1 (see Material and methods). The thermograms of the Cu-Ni-Ti 27 and Cu-Ni-Ti 35 products appear more alike than the 35 and 40 products, regardless of whether heating or cooling. Insofar as individual runs are concerned (Table II, M and A entries), the 3 Cu-Ni-Ti products repeatedly
Kusy and Whitley 233
American Journal of Orthodontics and Dentofacial Orthopedics Volume 131, Number 2
Table II.
Transition temperatures of products with first-order transformations via DSC Properties on heating
Number
AS (°C)
1 2 3
Cu-Ni-Ti 27
1 2 3
Cu-Ni-Ti 35
1 2 3
Cu-Ni-Ti 40
1 2 3
Orthonol
6.1 6.4 7.1 6.5 ⫾ 0.5§ 9.1 10.1 10.2 9.8 ⫾ 0.6 22.4 21.0 20.2 21.2 ⫾ 1.1 10.8 11.4 11.5 11.2 ⫾ 0.4
RS* (°C)
30.8 27.6 27.6 28.7 ⫾ 1.8
†
RF (°C)
35.7 30.2 30.1 32.0 ⫾ 3.2
Properties on cooling
AF (°C)
⌬HH ‡ (cal/g)
MS (°C)
MF (°C)
⌬HC ‡ (cal/g)
28.9 29.3 29.7 29.3 ⫾ 0.4 31.3 31.4 31.4 31.4 ⫾ 0.1 37.7 37.1 37.2 37.3 ⫾ 0.3 32.2 32.2 32.3 32.2 ⫾ 0.1
2.42 2.49 2.49 2.47 ⫾ 0.04 2.86 2.84 2.88 2.86 ⫾ 0.02 3.15 3.17 3.23 3.18 ⫾ 0.04 0.57 0.55 0.57 0.56 ⫾ 0.01
11.7 12.1 11.8 11.9 ⫾ 0.2 13.0 13.0 13.1 13.0 ⫾ 0.1 19.9 20.6 20.2 20.2 ⫾ 0.4 29.5 29.4 29.5 29.5 ⫾ 0.1
⫺15.4 ⫺14.3 ⫺14.0 ⫺14.6 ⫾ 0.7 ⫺9.9 ⫺10.8 ⫺10.7 ⫺10.5 ⫾ 0.5 5.7 4.5 4.7 5.0 ⫾ 0.6 8.8 8.5 9.1 8.8 ⫾ 0.3
2.78 2.77 2.77 2.77 ⫾ 0.01 3.22 3.23 3.23 3.23 ⫾ 0.01 3.39 3.48 3.49 3.46 ⫾ 0.05 0.57 0.57 0.57 0.57 ⫾ 0.00
See Fig 1 for definitions of AS, AF, ⌬HH, MS, MF, and ⌬HC. *Temperature of transformation from martensite to R-phase. † Temperature of transformation from R-phase to austenite. ‡ Where 1 cal/g ⫽ 4.18 Joule/g. § Mean ⫾ 1 SD.
suggest that the ends of the endothermic peaks on heating (ie, the temperature at which the thermogram returns to its baseline value on heating, AF) correspond to the manufacturer’s designations as follows: the 27 product vs 29.3°C, the 35 product vs 31.4°C, and the 40 product vs 37.3°C (see also references 9, 28, and 29). Overall, enthalpy changes for the 3 Cu-Ni-Ti and the Orthonol products (Table II, ⌬H entries) averaged about 3.00 and 0.57 calories per gram, respectively. For each Cu-Ni-Ti product, the magnitude of ⌬H increased as the transition temperature increased from the 27 to the 40 products, independent of heating or cooling. However, the ⌬HC values were always about 10% more than the values associated with ⌬HH. Overall, these observations agree with the results reported by McCoy30 and detailed in Orthodontic Materials: Scientific and Clinical Aspects.16 The means and standard deviations of those references were somewhat more scattered than our results. Table III summarizes the influence of frequency on the transition temperatures of the DMA technique. Unlike polymers, in which each decade of frequency shifts a transition by about 7C°,31 here the differences were only about 1C°. Consequently, only the lowest frequency needs to be further investigated for these intermetallic Ni-Ti alloys, although the mean values are just as informative, also. The DMA results at 0.1 Hz (Fig 5) corroborated
static tests on a tensile or bending machine.1-3 Specifically on heating, the storage modulus of the stainless steel alloy, Standard Rectangular, equaled 183 GPa at 20°C. The modulus of the beta-titanium alloy, TMA, equaled 64 GPa at 20°C. For the Ni-Ti alloy, Nitinol Classic, the modulus monotonically increased by 0.12 GPa per C° and equaled 41 GPa at 20°C. The only real surprise was the Orthonol product, which clearly exhibited an M ¡ A phase transformation as its modulus increased from 38 to 67 GPa. On cooling, the moduli values of this product generally replicated the heating traces, except for the 5C° to 10C° thermal delay that occurred between AS and MF as the transformation from A ¡ M started. When each Cu-Ni-Ti product was tested at 0.1 Hz, the values of ⌬T between the cooling and heating scans were 24.3°C and 23.3°C for Cu-Ni-Ti 27 and Cu-Ni-Ti 35, respectively, and 26.6°C for Cu-Ni-Ti 40 (Table III). When the cooling or heating scans of the Cu-Ni-Ti products were compared (Fig 6), the moduli of the low-temperature transformation ranged from 20 to 35 GPa, and the high-temperature transformation ranged from 60 to 70 GPa. Although the transformation region of Cu-Ni-Ti 27 and Cu-Ni-Ti 40 showed about a 10C° differential on heating and about a 7.5C° differential on cooling, the transformation region of Cu-Ni-Ti 35 showed a low start temperature (AS) that was most like Cu-Ni-Ti 27 and a high finish temperature (AF) that
234 Kusy and Whitley
Table III.
American Journal of Orthodontics and Dentofacial Orthopedics February 2007
Transition temperatures of products with first-order transformations via DMA Properties on heating
Properties on cooling
Products
AS (°C)
AF (°C)
Min. Dynamic Force (g)*
MS (°C)
MF (°C)
Max. Dynamic Force (g)*
⌬T (C°)
0.1 1.0 10.0
Cu-Ni-Ti 27
0.1 1.0 10.0
Cu-Ni-Ti 35
0.1 1.0 10.0
Cu-Ni-Ti 40
0.1 1.0 10.0
Orthonol
18.2 19.8 19.4 19.1 ⫾ 0.8† 21.2 20.9 20.9 21.0 ⫾ 0.2 31.0 29.9 29.6 30.2 ⫾ 0.7 22.4 22.2 22.4 22.3 ⫾ 0.1
27.5 27.5 27.3 27.4 ⫾ 0.1 36.5 35.6 35.2 35.8 ⫾ 0.7 39.2 39.9 39.8 39.6 ⫾ 0.4 36.4 35.9 36.1 36.1 ⫾ 0.3
11.6 13.1 13.5 12.7 ⫾ 1.0 13.8 15.3 15.7 14.9 ⫾ 1.0 14.0 14.8 15.1 14.6 ⫾ 0.6 19.3 19.8 20.2 19.8 ⫾ 0.5
5.4 6.4 6.5 6.1 ⫾ 0.6 13.7 14.0 13.2 13.6 ⫾ 0.4 14.2 15.1 15.6 15.0 ⫾ 0.7 30.2 30.0 30.5 30.2 ⫾ 0.3
⫺6.3 ⫺6.8 ⫺6.5 ⫺6.5 ⫾ 0.3 ⫺1.4 ⫺1.0 ⫺1.3 ⫺1.2 ⫾ 0.2 4.0 3.1 2.7 3.3 ⫾ 0.7 13.5 13.4 13.4 13.4 ⫾ 0.1
38.5 38.5 38.6 38.5 ⫾ 0.1 40.9 41.0 41.0 41.0 ⫾ 0.1 39.5 39.5 39.5 39.5 ⫾ 0.0 41.4 41.5 41.4 41.4 ⫾ 0.1
24.0 24.5 24.3 24.3 ⫾ 0.3 23.8 23.1 23.1 23.3 ⫾ 0.4 26.6 26.6 26.5 26.6 ⫾ 0.1 9.5 9.2 9.3 9.3 ⫾ 0.2
Frequency (Hz)
See Fig 2 for definitions of AS, AF, MS, MF, ⌬T, and minimum and maximum. *Where 10 g ⫽ 0.098 N. † Mean ⫾ 1 SD.
Fig 5. With DMA at 0.1 Hz, clearer differentiation occurs among materials shown in Fig 3.
was most like Cu-Ni-Ti 40. This combination of values maximized the range over which the transformational temperature of Cu-Ni-Ti 35 occurred, thereby decreasing the slope of its transformation region to the least among these 3 Cu-Ni-Ti products. When the minimum and maximum dynamic forces were compared, a factor of about 3 was found for the Cu-Ni-Ti products but only a factor of about 2 for the Orthonol product (Table III). These applied forces, used to maintain constant deflection during testing, are proportional to the ratios of the minimum and maxi-
Fig 6. With DMA at 0.1 Hz, further differentiation occurs among materials shown in Fig 4.
mum storage modulus values reported in Figure 6 and to the ratios of 2 to 4 reported by Kousbroek.23 DISCUSSION
At first glance, the connection between thermalmechanical characteristics and archwire mechanics appears remote. In reality, however, much insight can be gained about the true nature of commercial products— ie, how they might perform and whether they are a prudent investment. In this context, here are some
Kusy and Whitley 235
American Journal of Orthodontics and Dentofacial Orthopedics Volume 131, Number 2
Table IV.
Storage modulus (GPa) at 0.1 Hz at specific temperatures Storage modulus (GPa)*
Condition Heating
Cooling
Product
Ice cream @ 0°C
Room temperature @ 25°C
Oral temperature @ 34°C
Hot coffee @ 60°C
Cu-Ni-Ti 27 Cu-Ni-Ti 35 Cu-Ni-Ti 40 Orthonol Nitinol Classic TMA Standard Rectangular Cu-Ni-Ti 27 Cu-Ni-Ti 35 Cu-Ni-Ti 40 Orthonol Nitinol Classic TMA Standard Rectangular
26.4 28.2 29.1 39.7 40.2 66.5 187.4 41.0 30.5 30.2 36.3 40.7 65.0 186.8
44.9 31.2 24.6 37.3 41.3 63.7 182.4 59.0 60.0 64.2 56.9 42.9 64.3 183.8
58.6 53.3 40.2 55.9 42.1 66.3 181.2 61.7 62.6 64.4 65.4 44.2 64.1 183.2
63.6 64.4 65.1 67.4 46.0 63.6 179.2 65.9 66.6 67.5 69.9 48.1 63.3 180.1
*Determined from databases used to construct Figs 5 and 6.
examples of how evidence-based science complements experience-based practice. When Orthonol was first offered, this alloy behaved like another martensitic-stabilized alloy, Nitinol Classic. When heat flow is measured (Fig 3), a reversible transformation is suggested; this is better seen when the storage modulus is measured (Fig 5). Here one observes that Orthonol has biomechanical characteristics that mimic Nitinol Classic in the low-temperature region (T ⱕ10°C) but mimic TMA in the high-temperature region (T ⱖ40°C). Moreover, when the enthalpy is measured and compared with Cu-Ni-Ti wires (Table II, ⌬HC and ⌬HH entries), a value of 0.56 is obtained, which is only 20% that of a Cu-Ni-Ti wire. From these thermal-mechanical measurements (Figs 3 and 5, Table II), we adduce that Orthonol is about 20% thermoelastic active martensite, with the rest stable passive martensite. Meta-stable behavior is also observed in Ni-Ti archwires. In this context (Fig 4), Cu-Ni-Ti 40 suggests the presence of a meta-stable R-phase on heating. This does not occur on cooling, but it does recur when second and third heating cycles are run. From a clinician’s perspective, however, this intermediate transformation has no effect on the stiffness of this wire, as shown in the storage modulus data (Fig 6). Figure 6 broaches a clinical question of some importance. In a specialty with so many alternative products, are 3 Cu-Ni-Ti wires really necessary? The midpoints of the transformations follow expectations during heating or cooling. But, clinically speaking, the differences between Cu-Ni-Ti 27 and Cu-Ni-Ti 35 from
AS to AF and from MS to MF are only a few degrees (Table II). One advantage of Cu-Ni-Ti wires reportedly involves their thermal-activation characteristics. It is theorized that the ambient temperatures of various foods can activate and reactivate these wires. To investigate this phenomenon, we determined the stiffnesses from the averaged data (eg, Figs 5 and 6) at 4 salient temperatures9,11,13,14,32: ice cream (0°C), room temperature (25°C), oral temperature (34°C), and hot coffee (60°C). For such a thermal cycle, the following 4 cases can be considered (Table IV), all of which are influenced by hysteresis. Hot coffee and Cu-Ni-Ti 27: using this mode and the heating data, the practitioner takes the wire at 25°C (modulus ⫽ 44.9 GPa) and places it in the oral cavity at 34°C (modulus ⫽ 58.6 GPa). The patient drinks hot coffee at 60°C and raises the modulus to 63.6 GPa. As a consequence, the force per unit of deactivation is immediately increased by more than 40%. After the coffee is consumed, the cooling data show that the modulus (61.7 GPa) remains nearly constant down to oral temperature. Only when the temperature is substantially reduced below room temperature to about 0°C does the modulus drop to 41.0 GPa, a value below its starting modulus (Table IV, Fig 6). Hot coffee and Cu-Ni-Ti 40: here the practitioner takes a wire out of the package at 25°C (modulus ⫽ 24.6 GPa). At this point, this wire is more compliant than the other Cu-Ni-Ti wires (Table IV). When placed in the oral cavity at 34°C, the modulus rises somewhat (40.2 GPa), although it is still less than either Cu-Ni-Ti
236 Kusy and Whitley
27 or Cu-Ni-Ti 35 at 34°C. When the hot coffee is consumed, the modulus rises to 65.1 GPa, the same value found for all Cu-Ni-Ti wires. After cooling to oral temperature (34C°), the modulus (64.4 GPa) is very much like the 61.7 and 62.6 values that were found for Cu-Ni-Ti 27 and Cu-Ni-Ti 35, respectively. Ice cream and Cu-Ni-Ti 27: once again, the practitioner starts with the as-received wire at 25°C (modulus ⫽ 44.9 GPa). After the modulus increases to 58.6 GPa in the oral cavity, the patient eats a bowl of ice cream. This cooling reduces the modulus to as little as 41.0 GPa, which is very similar to that of the as-received product. Ice cream and Cu-Ni-Ti 40: in this case, the starting point of the as-received product has the lowest modulus (24.6 GPa). Once placed in the patient at 34°C, however, the modulus rises to 40.2 GPa. After drinking a cup of coffee during a meal, the patient consumes a bowl of ice cream for dessert. Now, the modulus drops to 30.2 GPa, which is more than the as-received wire (24.6 GPa) but only 75% the value at 34°C (40.2 GPa). Three more observations should be made with regard to Table IV— one that relates Orthonol to the Cu-Ni-Ti wires; another that considers Nitinol Classic, TMA, and Standard Rectangular wires relative to CuNi-Ti wires; and a final comment that uses Standard Rectangular wires to illustrate how far the materials of the specialty have progressed. First, over the entire scenario outlined in the 4 cases above, the Orthonol product transitions were comparable with the Cu-Ni-Ti 27 and Cu-Ni-Ti 35 alloys. For example, when presented in the order—Cu-Ni-Ti 27, Orthonol, and Cu-Ni-Ti 35— on heating, the moduli equaled 58.6, 55.9, and 53.3 GPa at 34°C and 63.6, 67.4, and 64.4 GPa at 60°C; on cooling, the moduli equaled 61.7, 65.4, and 62.6 GPa at 34°C. Second, the moduli of Nitinol Classic, TMA, and Standard Rectangular are rather invariant with temperature. Furthermore, Nitinol Classic has a modulus that is typical for stabilized martensite, but TMA appears most like the austenitic phase of Cu-Ni-Ti alloys. Finally, the Standard Rectangular product exemplifies how far clinical practice has progressed in its quest for light continuous forces—ie, with regard to the force per unit activation via the moduli, from a high of 183 GPa at 34°C to a low of 40.2 GPa for Cu-Ni-Ti 40. CONCLUSIONS
By using 2 techniques, the AF measurements of the Cu-Ni-Ti 27, 35, and 40 products equaled 29.3°C, 31.4°C, and 37.3°C by DSC and 27.4°C, 35.8°C, and 39.6°C by DMA. The latter technique is favored
American Journal of Orthodontics and Dentofacial Orthopedics February 2007
because it can simulate clinical scenarios and provide both thermal and mechanical data. The Orthonol product and the 3 Cu-Ni-Ti products had the stiffness qualities of Nitinol Classic at low temperatures and the stiffness qualities of TMA at high temperatures. This permits the wire to be more readily engaged with less patient discomfort because its stiffness is substantially less than the activated state at high temperature. “Does the transition temperature of Cu-Ni-Ti archwires affect the amount of tooth movement during alignment?” 33 If tooth movement is assumed to occur mainly when wires are activated by high-temperature excursions, our results show no difference between the alloys at high temperatures and little difference after cooling to oral temperature. The main difference occurs during engagement, in which Cu-Ni-Ti 40 provides the least force, and Cu-Ni-Ti 27 provides the most force.28,29 We thank the manufacturers for donating the materials for this study.
REFERENCES 1. Burstone CJ, Goldberg AJ. Beta titanium: a new orthodontic alloy. Am J Orthod 1980;77:121-32. 2. Kusy RP, Stush AM. Geometric and material parameters of a nickel-titanium and a beta titanium orthodontic arch wire alloy. Dent Mater 1987;3:207-17. 3. Kusy RP, Dilley GJ, Whitley JQ. Mechanical properties of stainless steel orthodontic archwires. Clin Mater 1988;33:41-59. 4. Rucker BK, Kusy RP. Elastic properties of alternative versus single-stranded leveling archwires. Am J Orthod Dentofacial Orthop 2002;122;528-41. 5. Rucker BK, Kusy RP. Elastic flexural properties of multistranded stainless steel versus conventional nickel-titanium archwires. Angle Orthod 2002;72:302-9. 6. Kusy RP, Greenberg AR. Effects of composition and cross section on the elastic properties of orthodontic wires. Angle Orthod 1981;51:325-41. 7. Kusy RP. Orthodontic biomaterials: from the past to the present. Angle Orthod 2002;72:501-12. 8. Kusy RP, Wilson TW. Dynamic mechanical properties of straight titanium alloy archwires. Dent Mater 1990;6:228-36. 9. Santoro M, Beshers DN. Nickel-titanium alloys: stress-related temperature transitional range. Am J Orthod Dentofacial Orthop 2000;118:685-92. 10. Es-Souni M, Es-Souni M, Brandies HF. On the transformation behaviour, mechanical properties and biocompatibility of two NiTi-based shape memory alloys: NiTi42 and NiTi42Cu7. Biomaterials 2001;22:2153-61. 11. Iijima M, Ohno H, Kawashima I, Endo K, Mizoguchi I. Mechanical behavior at different temperatures and stresses for superelastic nickel-titanium orthodontic wires having different transformation temperatures. Dent Mater 2002;18:88-93. 12. Brantley WA, Iijima M, Grentzer TH. Temperature-modulated DSC provides new insight about nickel-titanium wire transformations. Am J Orthod Dentofacial Orthop 2003;124:387-94.
Kusy and Whitley 237
American Journal of Orthodontics and Dentofacial Orthopedics Volume 131, Number 2
13. Kusy RP, Schafer DL. Rheology of stimulated whole saliva in a typical pre-orthodontic sample population. J Mater Sic: Mater Med 1995;6:385-9. 14. Gerlin A. How hot do you like it? The Wall Street Journal; Sept 1, 1994. Available from: http://www.vanfirm.com/mcdonaldscoffee-lawsuit.htm Accessed on March 3, 2005. 15. Wunderlich B. Thermal analysis. New York: Academic Press; 1990. p. 219-80, 350-62. 16. Brantley WA, Eliades T. Orthodontic materials: scientific and clinical aspects. Stuttgart, Germany: Thieme; 2001. p. 78-97. 17. Murayama T. Materials science monographs, 1: dynamic mechanical analysis of polymeric material. New York: Elsevier Scientific; 1978. p. 22. 18. Collings EW. The physical metallurgy of titanium alloys. Metals Park, Ohio: American Society for Metals; 1984. p. 148-62. 19. Donachie MJ Jr, editor. Titanium: a technical guide. Metals Park, Ohio: ASM International; 1988. p. 23-5. 20. Kusy RP. A review of contemporary archwires: their properties and characteristics. Angle Orthod 1997;67:197-208. 21. Hodgson DE, Ming HW, Biermann RJ. Shape memory alloys. Available from: http://www.jmmedical.com/html/_shape_memory_ alloys_.html. Accessed March, 3, 2005. 22. Gil FJ, Planell JA. Effect of copper addition on the superelastic behavior of Ni-Ti shape memory alloys for orthodontic applications. J Biomed Mater Res (Appl Biomater) 1999;48:682-8. 23. Kousbroek R. Shape memory alloys. In: Ducheyne P, Hastings GW, editors. Metal and ceramic biomaterials, volume II: strength and surface. Boca Raton, Fla: CRC Press; 1984. p. 65-7. 24. Gil FJ, Planell JA. Superelastic behavior of NiTiCu shape memory alloys for orthodontic applications. In: Advances in
25.
26. 27.
28.
29.
30.
31. 32.
33.
science and technology, vol 28 (Materials in clinical applications). Florence, Italy: Techna; 1999, p. 119-26. Meling TR, Ødegaard J. The effect of short-term temperature changes on superelastic nickel-titanium archwires activated in orthodontic bending. Am J Orthod Dentofacial Orthop 2001;119: 263-73. Callister WD Jr. Materials science and engineering: an introduction. 2nd ed. 1991. Wiley; New York. p. 690 – 6. Bradley TG, Brantley WA, Culbertson BM. Differential scanning calorimetry (DSC) analyses of superelastic and nonsuperelastic nickel-titanium orthodontic wires. Am J Orthod Dentofacial Orthop 1996;109:589-97. Filleul MP, Jordan L. Torsional properties of Ni-Ti and copper Ni-Ti wires: the effect of temperature on physical properties. Eur J Orthod 1997;19:637-46. Wilkinson PD, Dysart PS, Hood JAA, Herbison GP. Load-deflection characteristics of superelastic nickel-titanium orthodontic wires. Am J Orthod Dentofacial Orthop. 2002;121:483-95. McCoy BP. Comparison of compositions and differential scanning calorimetric analyses of the new copper-Ni-Ti wires with existing Ni-Ti orthodontic wires [thesis]. Columbus: Ohio State University; 1996. Nielsen LE. Mechanical properties of polymers and composites, volume 1. New York: Marcel Dekker; 1974. p. 143-61. Fischer-Brandies H, Es-Souni M, Kock N, Raetzke K, Bock O. Transformation behavior, chemical composition, surface topography and bending properties of five selected 0.016“ ⫻ 0.022” NiTi archwires. J Orofac Orthop 2003;64:88-99. Dalstra M, Melsen B. Does the transition temperature of Cu-NiTi archwires affect the amount of tooth movement during alignment? Orthod Craniofac Res 2004;7:21-5.
Thermal and mechanical characteristics of stainless steel, titanium-molybdenum, and nickel-titanium archwires Robert P. Kusya and John Q. Whitleyb Chapel Hill, NC Introduction: In recent years, nickel-titanium (Ni-Ti) archwires have been developed that undergo thermal transitions. Before the practitioner can fully utilize these products, the effect of those transitions within the clinical application must be understood. Methods: The transitional temperatures and mechanical stiffnesses of 3 archwire alloys—stainless steel, beta-titanium, and Ni-Ti—were investigated were for 7 products. Among the nickel-titanium alloys, 2 were thought to represent classic Ni-Ti products and 3 copper (Cu)-Ni-Ti products. By using 2 techniques, differential scanning calorimetry to measure heat flow and dynamic mechanical analysis to measure storage modulus, transition temperatures were evaluated from ⫺30°C to ⫹80°C. Results: With regard to the first technique, no transitions were observed for the stainless steel alloy, the beta-titanium alloy, and 1 of the 2 classic Ni-Ti products. For the other classic Ni-Ti product, however, a martensitic-austenitic transition was suggested on heating, and a reverse transformation was suggested on cooling. As expected, the Cu-Ni-Ti 27, 35, and 40 products manifested austenitic finish temperatures of 29.3°C, 31.4°C, and 37.3°C, respectively, as the enthalpy increased from 2.47 to 3.18 calories per gram. With regard to the second technique, the storage modulus at a low frequency of 0.1 Hz paralleled static mechanical tests for the stainless steel alloy (183 gigapascal [GPa]), the beta-titanium alloy (64 GPa), and the Nitinol Classic (3M Unitek, Monrovia, Calif) product that represented a stable martensitic phase (41 GPa). The remaining 4 Ni-Ti products generally varied from 20 to 35 GPa when the low-temperature or martensitic phase was present and from 60 to 70 GPa after the high-temperature or austenitic phase had formed. Conclusions: From the clinical viewpoint, the Orthonol (Rocky Mountain Orthodontics, Denver, Colo), Cu-Ni-Ti 27, Cu-Ni-Ti 35, and Cu-Ni-Ti 40 (SDS/Ormco, Glendora, Calif) products increased at least twofold in stiffness as temperature increased, best emulating the stiffness of Nitinol Classic below the transformational temperature and the stiffness of TMA (SDS/Ormco, Glendora, Calif) above the transformational temperature. Of the 3 Cu-Ni-Ti products, the least differences were found between Cu-Ni-Ti 27 and Cu-Ni-Ti 35, thereby questioning the justification for 3 similar products. (Am J Orthod Dentofacial Orthop 2007;131: 229-37)
P
revious investigators measured the tensile and flexure characteristics of archwires with differing outcomes, which depended on specific instrumentation, test methodology, and wire geometry.1-5 For example, stainless steel values varied from 22 ⫻ 106 to 30 ⫻ 106 psi,1,3 which in Systeme Internationale units corresponded to 150 to 200 gigapascal (GPa). Such differences profoundly influence the strength-torange ratio and consequently change the resilience by From the University of North Carolina, Chapel Hill. a Department of Orthodontics and Dental Research Center, School of Dentistry. b Dental Research Center, School of Dentistry. Supported in part by NIH grant 5 RO1 DE13201-03. Reprint requests to: Robert P. Kusy, University of North Carolina, DRC Building 210H, Room 313, CB#7455, Chapel Hill, NC 27599-7455; e-mail, [email protected]. Submitted, March 2005; revised and accepted, May 2005. 0889-5406/$32.00 Copyright © 2007 by the American Association of Orthodontists. doi:10.1016/j.ajodo.2005.05.054
the same proportions. Once either the working strength or its range is known along with the stiffness, the effectiveness of an archwire to move teeth is established.4,6,7 Primarily within the last 10 years, investigators have begun to characterize the transformational characteristics of archwires.8-12 Such studies are important for nickel-titanium (Ni-Ti) archwires whose reversible phase transformations occur from austenite to martensite, or sometimes an intermediate metastable R-phase occurs. Knowledge of the austenitic start and finish temperatures (on heating) and the martensitic start and finish temperatures (on cooling) establishes the stiffnesses of each archwire at ambient and oral-cavity conditions (34°C),13 and at the temperatures of ice cream (about 0°C) and hot coffee (about 60°C).14 Since the first commercial thermal analysis instrumentation in the 1960s, hardware and software capa229
230 Kusy and Whitley
bilities have grown so that today some 2 dozen techniques are available—2 of which are differential scanning calorimetry (DSC) and dynamic mechanical analysis (DMA). DSC measures heat flow as temperature changes.15 When changes occur from solid to liquid states (eg, the melting of ice) or, in the present case, from 1 phase to another in the same state (from martensite to austenite in the solid state), an endothermic (or exothermic) reaction occurs as a result of heating (or cooling). Thereby, the transitional temperatures and enthalpy changes associated with latent heat of transition can be measured.16 On the other hand, DMA measures the real component of the complex modulus—the storage modulus—as the temperature changes.17 When such a specimen is appropriately vibrated at 1 or more known frequencies, the response might not only lag behind but also become attenuated. These changes are most apparent as the temperature traverses a transformation such as a solid-state phase change that is found in some Ni-Ti products.18 In this article, the DSC and DMA techniques are used to determine both the phase transformations and the magnitude of storage moduli (a measure of wire stiffness) for 3 alloy groups. From these measurements, we show not only the moduli of invariant alloys but also the temperatures, enthalpies, and moduli of alloys undergoing solid-state phase transformations. Specifically, Ni-Ti products can exhibit storage moduli that vary with temperature by a factor of 2, when only small changes in composition and temperature occur. Therefore, Ni-Ti products should be selected on the basis of their properties and clinical applications—not on the basis of their costs and vendors. MATERIAL AND METHODS
Seven archwire products comprising 3 alloy groups were studied (Table I): a traditional austenitic stainless steel, the first-marketed titanium-molybdenum binary called beta-titanium, and 5 Ni-Ti alloys. Of the 5 Ni-Ti alloys, 2 were thought to be stabilized martensitic alloys in which processing prevented further transformations, and 3 were thought to be active alloys in which the amounts of copper (Cu) and chromium19,20 influenced the transformations from austenite to martensite on cooling (A ¡ M) and martensite to austenite on heating (M ¡ A).21,22 When both transformations were possible, the alloy was said to be reversibly thermoelastic, and the notation appears either as M ↔ A or, equivalently, as A ↔ M. In standard DSC mode, specimens (10-11 mg) of wire products were scanned 3 times with a differential scanning calorimeter (Q100, TA Instruments, New Castle, Del) and an attached refrigerated cooling sys-
American Journal of Orthodontics and Dentofacial Orthopedics February 2007
Table I.
Materials evaluated
Product
Alloy
Standard Rectangular TMA Nitinol Classic Orthonol
Stainless steel
Cu-Ni-Ti 27 Cu-Ni-Ti 35 Cu-Ni-Ti 40
Nickel-titanium Nickel-titanium Nickel-titanium
Beta-titanium Nickel-titanium Nickel-titanium
Constituents
Manufacturer
Fe, Cr, Ni, C
American Orthodontics Ti, Mo, Zr, Sn SDS/Ormco Ni, Ti 3M Unitek Ni, Ti Rocky Mountain Orthodontics Ni, Ti, Cu, Cr SDS/Ormco Ni, Ti, Cu, Cr SDS/Ormco Ni, Ti, Cu, Cr SDS/Ormco
All wire specimens were nominally 0.016 ⫻ 0.022 in. Mo, Molybdenum; Zr, zirconium; Sn, tin; Cr, chromium; Fe, iron; C, carbon; Ni, nickel; Cu, copper; Sn, tin; Ti, titanium.
tem (Fig 1, A). The instrument was calibrated with matched sapphire disks and then was run with indium. All specimens were scanned from ⫺30°C to 60°C under an atmosphere of ultrahigh purity nitrogen. Immediately after this temperature ramp, the specimens were scanned from 60 to ⫺30°C at the same rate. When both the specimen and the reference pans were heated at the programmable rate of 5C° per minute in the temperature-controlled chamber (Fig 1, B), changes in heat flow occurred when a phase transition happened. (In this article, there is an important difference between 5°C and 5C°. The former refers to a specific temperature on the Celsius scale; the latter indicates the magnitude of a thermal change that is independent of the actual Celsius temperature.) When the phase transition is a solid-solid transformation, an endothermic absorption peak is observed on heating that begins at the austenite start (AS) temperature and ends at the austenite finish (AF) temperature.23 This peak appears like other first-order transformations— eg, melting amd boiling peaks. After cooling at the same programmable rate, the transformation occurs somewhat later, and the phase change appears supercooled. That is, the transformation of A ¡ M occurs at the martensite start (MS) temperature, and the completion occurs at the martensite finish (MF) temperature (Fig 1, C ). After we measured these 4 temperatures, we calculated the endothermic peaks on heating and the exothermic peaks on cooling. These peaks are proportional to the enthalpies of heating (⌬HH) and cooling (⌬HC), respectively, and were calculated by using the sigmoidal tangent algorithm of Universal Analysis 2002 (TA Instruments). By using a 5-mm span, wire specimens were scanned in the 3-point bending fixture of a dynamic mechanical analyzer (2980, TA Instruments) outfitted with a liquid nitrogen gas cooling accessory (Fig 2, A). After preloading with 5 g at an amplitude of 0.0005 cm,
American Journal of Orthodontics and Dentofacial Orthopedics Volume 131, Number 2
Kusy and Whitley 231
Fig 1. DSC: A, Loading platform with centrally located reference pan and specimen pan; B, schematic of temperature-controlled chamber in vertical plane; C, typical heating and cooling thermograms defining 6 parameters of importance.
Fig 2. DMA: A, Three-point bending fixture; B, schematic of temperature-controlled chamber and its hardware in vertical plane; C, typical heating and cooling thermograms defining 7 parameters of importance.
the specimens were scanned by using frictionless air bearings and a drive motor from ⫺20° to 80°C at a rate of 1C° per minute. This was followed by a scan from 80° to ⫺30°C at the same rate in a temperature-
controlled chamber (Fig 2, B). The responses from 3 frequencies (0.1, 1.0, and 10 Hz) were sequentially measured at each temperature with an optical encoder. The lowest frequency best mimicked the static test
232 Kusy and Whitley
conditions of conventional tensile or bending tests. By using the same phase notation as the DSC tester, each storage modulus (which is directly proportional to the dynamic force applied) showed a similar relative delay of the cooling curve compared with its heating curve (Fig 2, C). This shift resulted in a hysteresis loop,22-25 which (as in hard vs soft magnets26) indicates the ease of the transformation. The low temperature phase (M) at temperature (T) ⬍MF has a lower storage modulus than the high temperature phase (A) at T ⬎AF. Consequently, on heating, the dynamic force is at its minimum for the martensitic phase as AS is approached and at its maximum for the austenitic phase once AF has been exceeded. On subsequent cooling, the A phase persists down to MS, but transformation to all M phase is not complete until MF. The horizontal magnitude associated with each hysteresis loop (⌬T) was calculated by first determining the storage modulus at the midpoint between AS and AF of the heating curve. Then, by extrapolating this modulus onto the cooling curve, the difference between the temperature of the heating and cooling traces, ⌬T, was obtained.
American Journal of Orthodontics and Dentofacial Orthopedics February 2007
Fig 3. Comparison of DSC thermograms of 4 archwires for which no transitions were expected: (top to bottom) Standard Rectangular, TMA, Nitinol Classic, and Orthonol (see Table I).
Statistical analysis
For both the DSC and DMA techniques, the means ⫾ 1 SD were determined for transition temperatures of all products that had first-order transformations. In both techniques, specific values were averaged: for DSC, representing the 3 runs; for DMA, these were the 3 frequencies. RESULTS
Figure 3 shows the DSC results, from top to bottom, of the heating and cooling thermograms for stainless steel, beta-titanium, and 2 Ni-Ti alloys, respectively. As expected, the Standard Rectangular (Am Orthodontics, Sheboygan, Wisc), TMA (SDS/Ormco, Glendora, Calif), and the Nitinol Classic (3M Unitek, Monrovia, Calif) products showed no transitions in the temperature regime of the oral cavity. Contrary to earlier beliefs and possibly because of a change in this Ni-Ti alloy, the Orthonol (Rocky Mountain Orthodontics, Denver, Colo) product had a small endothermic (or exothermic) peak on heating (or cooling).8 Whether this represents an M ¡ A transformation on heating and an A ¡ M transformation on cooling will be verified by the DMA results. Figure 4 details the 3 Cu-Ni-Ti products, which the manufacturer designates by their transitions of 27°C, 35°C, and 40°C (SDS/Ormco). When heated, all CuNi-Ti products had endothermic reactions in the order that their thermal monikers would imply. Two distinct transformations occurred for Cu-Ni-Ti 40, the interme-
Fig 4. Comparison of DSC thermograms in which at least 1 transition occurs on heating and cooling: (top to bottom) Cu-Ni-Ti 27, Cu-Ni-Ti 35, and Cu-Ni-Ti 40 (see Table I).
diate R-phase27 followed by the high-temperature or austenitic phase. When cooled, all had the appearance shown in Figure 1 (see Material and methods). The thermograms of the Cu-Ni-Ti 27 and Cu-Ni-Ti 35 products appear more alike than the 35 and 40 products, regardless of whether heating or cooling. Insofar as individual runs are concerned (Table II, M and A entries), the 3 Cu-Ni-Ti products repeatedly
Kusy and Whitley 233
American Journal of Orthodontics and Dentofacial Orthopedics Volume 131, Number 2
Table II.
Transition temperatures of products with first-order transformations via DSC Properties on heating
Number
AS (°C)
1 2 3
Cu-Ni-Ti 27
1 2 3
Cu-Ni-Ti 35
1 2 3
Cu-Ni-Ti 40
1 2 3
Orthonol
6.1 6.4 7.1 6.5 ⫾ 0.5§ 9.1 10.1 10.2 9.8 ⫾ 0.6 22.4 21.0 20.2 21.2 ⫾ 1.1 10.8 11.4 11.5 11.2 ⫾ 0.4
RS* (°C)
30.8 27.6 27.6 28.7 ⫾ 1.8
†
RF (°C)
35.7 30.2 30.1 32.0 ⫾ 3.2
Properties on cooling
AF (°C)
⌬HH ‡ (cal/g)
MS (°C)
MF (°C)
⌬HC ‡ (cal/g)
28.9 29.3 29.7 29.3 ⫾ 0.4 31.3 31.4 31.4 31.4 ⫾ 0.1 37.7 37.1 37.2 37.3 ⫾ 0.3 32.2 32.2 32.3 32.2 ⫾ 0.1
2.42 2.49 2.49 2.47 ⫾ 0.04 2.86 2.84 2.88 2.86 ⫾ 0.02 3.15 3.17 3.23 3.18 ⫾ 0.04 0.57 0.55 0.57 0.56 ⫾ 0.01
11.7 12.1 11.8 11.9 ⫾ 0.2 13.0 13.0 13.1 13.0 ⫾ 0.1 19.9 20.6 20.2 20.2 ⫾ 0.4 29.5 29.4 29.5 29.5 ⫾ 0.1
⫺15.4 ⫺14.3 ⫺14.0 ⫺14.6 ⫾ 0.7 ⫺9.9 ⫺10.8 ⫺10.7 ⫺10.5 ⫾ 0.5 5.7 4.5 4.7 5.0 ⫾ 0.6 8.8 8.5 9.1 8.8 ⫾ 0.3
2.78 2.77 2.77 2.77 ⫾ 0.01 3.22 3.23 3.23 3.23 ⫾ 0.01 3.39 3.48 3.49 3.46 ⫾ 0.05 0.57 0.57 0.57 0.57 ⫾ 0.00
See Fig 1 for definitions of AS, AF, ⌬HH, MS, MF, and ⌬HC. *Temperature of transformation from martensite to R-phase. † Temperature of transformation from R-phase to austenite. ‡ Where 1 cal/g ⫽ 4.18 Joule/g. § Mean ⫾ 1 SD.
suggest that the ends of the endothermic peaks on heating (ie, the temperature at which the thermogram returns to its baseline value on heating, AF) correspond to the manufacturer’s designations as follows: the 27 product vs 29.3°C, the 35 product vs 31.4°C, and the 40 product vs 37.3°C (see also references 9, 28, and 29). Overall, enthalpy changes for the 3 Cu-Ni-Ti and the Orthonol products (Table II, ⌬H entries) averaged about 3.00 and 0.57 calories per gram, respectively. For each Cu-Ni-Ti product, the magnitude of ⌬H increased as the transition temperature increased from the 27 to the 40 products, independent of heating or cooling. However, the ⌬HC values were always about 10% more than the values associated with ⌬HH. Overall, these observations agree with the results reported by McCoy30 and detailed in Orthodontic Materials: Scientific and Clinical Aspects.16 The means and standard deviations of those references were somewhat more scattered than our results. Table III summarizes the influence of frequency on the transition temperatures of the DMA technique. Unlike polymers, in which each decade of frequency shifts a transition by about 7C°,31 here the differences were only about 1C°. Consequently, only the lowest frequency needs to be further investigated for these intermetallic Ni-Ti alloys, although the mean values are just as informative, also. The DMA results at 0.1 Hz (Fig 5) corroborated
static tests on a tensile or bending machine.1-3 Specifically on heating, the storage modulus of the stainless steel alloy, Standard Rectangular, equaled 183 GPa at 20°C. The modulus of the beta-titanium alloy, TMA, equaled 64 GPa at 20°C. For the Ni-Ti alloy, Nitinol Classic, the modulus monotonically increased by 0.12 GPa per C° and equaled 41 GPa at 20°C. The only real surprise was the Orthonol product, which clearly exhibited an M ¡ A phase transformation as its modulus increased from 38 to 67 GPa. On cooling, the moduli values of this product generally replicated the heating traces, except for the 5C° to 10C° thermal delay that occurred between AS and MF as the transformation from A ¡ M started. When each Cu-Ni-Ti product was tested at 0.1 Hz, the values of ⌬T between the cooling and heating scans were 24.3°C and 23.3°C for Cu-Ni-Ti 27 and Cu-Ni-Ti 35, respectively, and 26.6°C for Cu-Ni-Ti 40 (Table III). When the cooling or heating scans of the Cu-Ni-Ti products were compared (Fig 6), the moduli of the low-temperature transformation ranged from 20 to 35 GPa, and the high-temperature transformation ranged from 60 to 70 GPa. Although the transformation region of Cu-Ni-Ti 27 and Cu-Ni-Ti 40 showed about a 10C° differential on heating and about a 7.5C° differential on cooling, the transformation region of Cu-Ni-Ti 35 showed a low start temperature (AS) that was most like Cu-Ni-Ti 27 and a high finish temperature (AF) that
234 Kusy and Whitley
Table III.
American Journal of Orthodontics and Dentofacial Orthopedics February 2007
Transition temperatures of products with first-order transformations via DMA Properties on heating
Properties on cooling
Products
AS (°C)
AF (°C)
Min. Dynamic Force (g)*
MS (°C)
MF (°C)
Max. Dynamic Force (g)*
⌬T (C°)
0.1 1.0 10.0
Cu-Ni-Ti 27
0.1 1.0 10.0
Cu-Ni-Ti 35
0.1 1.0 10.0
Cu-Ni-Ti 40
0.1 1.0 10.0
Orthonol
18.2 19.8 19.4 19.1 ⫾ 0.8† 21.2 20.9 20.9 21.0 ⫾ 0.2 31.0 29.9 29.6 30.2 ⫾ 0.7 22.4 22.2 22.4 22.3 ⫾ 0.1
27.5 27.5 27.3 27.4 ⫾ 0.1 36.5 35.6 35.2 35.8 ⫾ 0.7 39.2 39.9 39.8 39.6 ⫾ 0.4 36.4 35.9 36.1 36.1 ⫾ 0.3
11.6 13.1 13.5 12.7 ⫾ 1.0 13.8 15.3 15.7 14.9 ⫾ 1.0 14.0 14.8 15.1 14.6 ⫾ 0.6 19.3 19.8 20.2 19.8 ⫾ 0.5
5.4 6.4 6.5 6.1 ⫾ 0.6 13.7 14.0 13.2 13.6 ⫾ 0.4 14.2 15.1 15.6 15.0 ⫾ 0.7 30.2 30.0 30.5 30.2 ⫾ 0.3
⫺6.3 ⫺6.8 ⫺6.5 ⫺6.5 ⫾ 0.3 ⫺1.4 ⫺1.0 ⫺1.3 ⫺1.2 ⫾ 0.2 4.0 3.1 2.7 3.3 ⫾ 0.7 13.5 13.4 13.4 13.4 ⫾ 0.1
38.5 38.5 38.6 38.5 ⫾ 0.1 40.9 41.0 41.0 41.0 ⫾ 0.1 39.5 39.5 39.5 39.5 ⫾ 0.0 41.4 41.5 41.4 41.4 ⫾ 0.1
24.0 24.5 24.3 24.3 ⫾ 0.3 23.8 23.1 23.1 23.3 ⫾ 0.4 26.6 26.6 26.5 26.6 ⫾ 0.1 9.5 9.2 9.3 9.3 ⫾ 0.2
Frequency (Hz)
See Fig 2 for definitions of AS, AF, MS, MF, ⌬T, and minimum and maximum. *Where 10 g ⫽ 0.098 N. † Mean ⫾ 1 SD.
Fig 5. With DMA at 0.1 Hz, clearer differentiation occurs among materials shown in Fig 3.
was most like Cu-Ni-Ti 40. This combination of values maximized the range over which the transformational temperature of Cu-Ni-Ti 35 occurred, thereby decreasing the slope of its transformation region to the least among these 3 Cu-Ni-Ti products. When the minimum and maximum dynamic forces were compared, a factor of about 3 was found for the Cu-Ni-Ti products but only a factor of about 2 for the Orthonol product (Table III). These applied forces, used to maintain constant deflection during testing, are proportional to the ratios of the minimum and maxi-
Fig 6. With DMA at 0.1 Hz, further differentiation occurs among materials shown in Fig 4.
mum storage modulus values reported in Figure 6 and to the ratios of 2 to 4 reported by Kousbroek.23 DISCUSSION
At first glance, the connection between thermalmechanical characteristics and archwire mechanics appears remote. In reality, however, much insight can be gained about the true nature of commercial products— ie, how they might perform and whether they are a prudent investment. In this context, here are some
Kusy and Whitley 235
American Journal of Orthodontics and Dentofacial Orthopedics Volume 131, Number 2
Table IV.
Storage modulus (GPa) at 0.1 Hz at specific temperatures Storage modulus (GPa)*
Condition Heating
Cooling
Product
Ice cream @ 0°C
Room temperature @ 25°C
Oral temperature @ 34°C
Hot coffee @ 60°C
Cu-Ni-Ti 27 Cu-Ni-Ti 35 Cu-Ni-Ti 40 Orthonol Nitinol Classic TMA Standard Rectangular Cu-Ni-Ti 27 Cu-Ni-Ti 35 Cu-Ni-Ti 40 Orthonol Nitinol Classic TMA Standard Rectangular
26.4 28.2 29.1 39.7 40.2 66.5 187.4 41.0 30.5 30.2 36.3 40.7 65.0 186.8
44.9 31.2 24.6 37.3 41.3 63.7 182.4 59.0 60.0 64.2 56.9 42.9 64.3 183.8
58.6 53.3 40.2 55.9 42.1 66.3 181.2 61.7 62.6 64.4 65.4 44.2 64.1 183.2
63.6 64.4 65.1 67.4 46.0 63.6 179.2 65.9 66.6 67.5 69.9 48.1 63.3 180.1
*Determined from databases used to construct Figs 5 and 6.
examples of how evidence-based science complements experience-based practice. When Orthonol was first offered, this alloy behaved like another martensitic-stabilized alloy, Nitinol Classic. When heat flow is measured (Fig 3), a reversible transformation is suggested; this is better seen when the storage modulus is measured (Fig 5). Here one observes that Orthonol has biomechanical characteristics that mimic Nitinol Classic in the low-temperature region (T ⱕ10°C) but mimic TMA in the high-temperature region (T ⱖ40°C). Moreover, when the enthalpy is measured and compared with Cu-Ni-Ti wires (Table II, ⌬HC and ⌬HH entries), a value of 0.56 is obtained, which is only 20% that of a Cu-Ni-Ti wire. From these thermal-mechanical measurements (Figs 3 and 5, Table II), we adduce that Orthonol is about 20% thermoelastic active martensite, with the rest stable passive martensite. Meta-stable behavior is also observed in Ni-Ti archwires. In this context (Fig 4), Cu-Ni-Ti 40 suggests the presence of a meta-stable R-phase on heating. This does not occur on cooling, but it does recur when second and third heating cycles are run. From a clinician’s perspective, however, this intermediate transformation has no effect on the stiffness of this wire, as shown in the storage modulus data (Fig 6). Figure 6 broaches a clinical question of some importance. In a specialty with so many alternative products, are 3 Cu-Ni-Ti wires really necessary? The midpoints of the transformations follow expectations during heating or cooling. But, clinically speaking, the differences between Cu-Ni-Ti 27 and Cu-Ni-Ti 35 from
AS to AF and from MS to MF are only a few degrees (Table II). One advantage of Cu-Ni-Ti wires reportedly involves their thermal-activation characteristics. It is theorized that the ambient temperatures of various foods can activate and reactivate these wires. To investigate this phenomenon, we determined the stiffnesses from the averaged data (eg, Figs 5 and 6) at 4 salient temperatures9,11,13,14,32: ice cream (0°C), room temperature (25°C), oral temperature (34°C), and hot coffee (60°C). For such a thermal cycle, the following 4 cases can be considered (Table IV), all of which are influenced by hysteresis. Hot coffee and Cu-Ni-Ti 27: using this mode and the heating data, the practitioner takes the wire at 25°C (modulus ⫽ 44.9 GPa) and places it in the oral cavity at 34°C (modulus ⫽ 58.6 GPa). The patient drinks hot coffee at 60°C and raises the modulus to 63.6 GPa. As a consequence, the force per unit of deactivation is immediately increased by more than 40%. After the coffee is consumed, the cooling data show that the modulus (61.7 GPa) remains nearly constant down to oral temperature. Only when the temperature is substantially reduced below room temperature to about 0°C does the modulus drop to 41.0 GPa, a value below its starting modulus (Table IV, Fig 6). Hot coffee and Cu-Ni-Ti 40: here the practitioner takes a wire out of the package at 25°C (modulus ⫽ 24.6 GPa). At this point, this wire is more compliant than the other Cu-Ni-Ti wires (Table IV). When placed in the oral cavity at 34°C, the modulus rises somewhat (40.2 GPa), although it is still less than either Cu-Ni-Ti
236 Kusy and Whitley
27 or Cu-Ni-Ti 35 at 34°C. When the hot coffee is consumed, the modulus rises to 65.1 GPa, the same value found for all Cu-Ni-Ti wires. After cooling to oral temperature (34C°), the modulus (64.4 GPa) is very much like the 61.7 and 62.6 values that were found for Cu-Ni-Ti 27 and Cu-Ni-Ti 35, respectively. Ice cream and Cu-Ni-Ti 27: once again, the practitioner starts with the as-received wire at 25°C (modulus ⫽ 44.9 GPa). After the modulus increases to 58.6 GPa in the oral cavity, the patient eats a bowl of ice cream. This cooling reduces the modulus to as little as 41.0 GPa, which is very similar to that of the as-received product. Ice cream and Cu-Ni-Ti 40: in this case, the starting point of the as-received product has the lowest modulus (24.6 GPa). Once placed in the patient at 34°C, however, the modulus rises to 40.2 GPa. After drinking a cup of coffee during a meal, the patient consumes a bowl of ice cream for dessert. Now, the modulus drops to 30.2 GPa, which is more than the as-received wire (24.6 GPa) but only 75% the value at 34°C (40.2 GPa). Three more observations should be made with regard to Table IV— one that relates Orthonol to the Cu-Ni-Ti wires; another that considers Nitinol Classic, TMA, and Standard Rectangular wires relative to CuNi-Ti wires; and a final comment that uses Standard Rectangular wires to illustrate how far the materials of the specialty have progressed. First, over the entire scenario outlined in the 4 cases above, the Orthonol product transitions were comparable with the Cu-Ni-Ti 27 and Cu-Ni-Ti 35 alloys. For example, when presented in the order—Cu-Ni-Ti 27, Orthonol, and Cu-Ni-Ti 35— on heating, the moduli equaled 58.6, 55.9, and 53.3 GPa at 34°C and 63.6, 67.4, and 64.4 GPa at 60°C; on cooling, the moduli equaled 61.7, 65.4, and 62.6 GPa at 34°C. Second, the moduli of Nitinol Classic, TMA, and Standard Rectangular are rather invariant with temperature. Furthermore, Nitinol Classic has a modulus that is typical for stabilized martensite, but TMA appears most like the austenitic phase of Cu-Ni-Ti alloys. Finally, the Standard Rectangular product exemplifies how far clinical practice has progressed in its quest for light continuous forces—ie, with regard to the force per unit activation via the moduli, from a high of 183 GPa at 34°C to a low of 40.2 GPa for Cu-Ni-Ti 40. CONCLUSIONS
By using 2 techniques, the AF measurements of the Cu-Ni-Ti 27, 35, and 40 products equaled 29.3°C, 31.4°C, and 37.3°C by DSC and 27.4°C, 35.8°C, and 39.6°C by DMA. The latter technique is favored
American Journal of Orthodontics and Dentofacial Orthopedics February 2007
because it can simulate clinical scenarios and provide both thermal and mechanical data. The Orthonol product and the 3 Cu-Ni-Ti products had the stiffness qualities of Nitinol Classic at low temperatures and the stiffness qualities of TMA at high temperatures. This permits the wire to be more readily engaged with less patient discomfort because its stiffness is substantially less than the activated state at high temperature. “Does the transition temperature of Cu-Ni-Ti archwires affect the amount of tooth movement during alignment?” 33 If tooth movement is assumed to occur mainly when wires are activated by high-temperature excursions, our results show no difference between the alloys at high temperatures and little difference after cooling to oral temperature. The main difference occurs during engagement, in which Cu-Ni-Ti 40 provides the least force, and Cu-Ni-Ti 27 provides the most force.28,29 We thank the manufacturers for donating the materials for this study.
REFERENCES 1. Burstone CJ, Goldberg AJ. Beta titanium: a new orthodontic alloy. Am J Orthod 1980;77:121-32. 2. Kusy RP, Stush AM. Geometric and material parameters of a nickel-titanium and a beta titanium orthodontic arch wire alloy. Dent Mater 1987;3:207-17. 3. Kusy RP, Dilley GJ, Whitley JQ. Mechanical properties of stainless steel orthodontic archwires. Clin Mater 1988;33:41-59. 4. Rucker BK, Kusy RP. Elastic properties of alternative versus single-stranded leveling archwires. Am J Orthod Dentofacial Orthop 2002;122;528-41. 5. Rucker BK, Kusy RP. Elastic flexural properties of multistranded stainless steel versus conventional nickel-titanium archwires. Angle Orthod 2002;72:302-9. 6. Kusy RP, Greenberg AR. Effects of composition and cross section on the elastic properties of orthodontic wires. Angle Orthod 1981;51:325-41. 7. Kusy RP. Orthodontic biomaterials: from the past to the present. Angle Orthod 2002;72:501-12. 8. Kusy RP, Wilson TW. Dynamic mechanical properties of straight titanium alloy archwires. Dent Mater 1990;6:228-36. 9. Santoro M, Beshers DN. Nickel-titanium alloys: stress-related temperature transitional range. Am J Orthod Dentofacial Orthop 2000;118:685-92. 10. Es-Souni M, Es-Souni M, Brandies HF. On the transformation behaviour, mechanical properties and biocompatibility of two NiTi-based shape memory alloys: NiTi42 and NiTi42Cu7. Biomaterials 2001;22:2153-61. 11. Iijima M, Ohno H, Kawashima I, Endo K, Mizoguchi I. Mechanical behavior at different temperatures and stresses for superelastic nickel-titanium orthodontic wires having different transformation temperatures. Dent Mater 2002;18:88-93. 12. Brantley WA, Iijima M, Grentzer TH. Temperature-modulated DSC provides new insight about nickel-titanium wire transformations. Am J Orthod Dentofacial Orthop 2003;124:387-94.
Kusy and Whitley 237
American Journal of Orthodontics and Dentofacial Orthopedics Volume 131, Number 2
13. Kusy RP, Schafer DL. Rheology of stimulated whole saliva in a typical pre-orthodontic sample population. J Mater Sic: Mater Med 1995;6:385-9. 14. Gerlin A. How hot do you like it? The Wall Street Journal; Sept 1, 1994. Available from: http://www.vanfirm.com/mcdonaldscoffee-lawsuit.htm Accessed on March 3, 2005. 15. Wunderlich B. Thermal analysis. New York: Academic Press; 1990. p. 219-80, 350-62. 16. Brantley WA, Eliades T. Orthodontic materials: scientific and clinical aspects. Stuttgart, Germany: Thieme; 2001. p. 78-97. 17. Murayama T. Materials science monographs, 1: dynamic mechanical analysis of polymeric material. New York: Elsevier Scientific; 1978. p. 22. 18. Collings EW. The physical metallurgy of titanium alloys. Metals Park, Ohio: American Society for Metals; 1984. p. 148-62. 19. Donachie MJ Jr, editor. Titanium: a technical guide. Metals Park, Ohio: ASM International; 1988. p. 23-5. 20. Kusy RP. A review of contemporary archwires: their properties and characteristics. Angle Orthod 1997;67:197-208. 21. Hodgson DE, Ming HW, Biermann RJ. Shape memory alloys. Available from: http://www.jmmedical.com/html/_shape_memory_ alloys_.html. Accessed March, 3, 2005. 22. Gil FJ, Planell JA. Effect of copper addition on the superelastic behavior of Ni-Ti shape memory alloys for orthodontic applications. J Biomed Mater Res (Appl Biomater) 1999;48:682-8. 23. Kousbroek R. Shape memory alloys. In: Ducheyne P, Hastings GW, editors. Metal and ceramic biomaterials, volume II: strength and surface. Boca Raton, Fla: CRC Press; 1984. p. 65-7. 24. Gil FJ, Planell JA. Superelastic behavior of NiTiCu shape memory alloys for orthodontic applications. In: Advances in
25.
26. 27.
28.
29.
30.
31. 32.
33.
science and technology, vol 28 (Materials in clinical applications). Florence, Italy: Techna; 1999, p. 119-26. Meling TR, Ødegaard J. The effect of short-term temperature changes on superelastic nickel-titanium archwires activated in orthodontic bending. Am J Orthod Dentofacial Orthop 2001;119: 263-73. Callister WD Jr. Materials science and engineering: an introduction. 2nd ed. 1991. Wiley; New York. p. 690 – 6. Bradley TG, Brantley WA, Culbertson BM. Differential scanning calorimetry (DSC) analyses of superelastic and nonsuperelastic nickel-titanium orthodontic wires. Am J Orthod Dentofacial Orthop 1996;109:589-97. Filleul MP, Jordan L. Torsional properties of Ni-Ti and copper Ni-Ti wires: the effect of temperature on physical properties. Eur J Orthod 1997;19:637-46. Wilkinson PD, Dysart PS, Hood JAA, Herbison GP. Load-deflection characteristics of superelastic nickel-titanium orthodontic wires. Am J Orthod Dentofacial Orthop. 2002;121:483-95. McCoy BP. Comparison of compositions and differential scanning calorimetric analyses of the new copper-Ni-Ti wires with existing Ni-Ti orthodontic wires [thesis]. Columbus: Ohio State University; 1996. Nielsen LE. Mechanical properties of polymers and composites, volume 1. New York: Marcel Dekker; 1974. p. 143-61. Fischer-Brandies H, Es-Souni M, Kock N, Raetzke K, Bock O. Transformation behavior, chemical composition, surface topography and bending properties of five selected 0.016“ ⫻ 0.022” NiTi archwires. J Orofac Orthop 2003;64:88-99. Dalstra M, Melsen B. Does the transition temperature of Cu-NiTi archwires affect the amount of tooth movement during alignment? Orthod Craniofac Res 2004;7:21-5.