P. Winiarski, A. Kłossowicz, W. Stęplewski, J. Wróblewski, A. Dziedzic; Electrical properties of thin-film resistors in a wide temperature range, Circuit World, vol.41(2015), p.116-120

Electrical properties of thin-film resistors in a wide temperature range Paweł Winiarski1 , Adam Kłossowicz1 , Jacek Wróble wski1 , Andrze j Dziedzic1 , Wojciech Stęplewski2 Wrocław University of Technology, Faculty of Microsystem Electronics and Photonics, Wybrzeże Wyspiańskiego 27, 50-370 Wrocław, Poland, e-mail: [email protected] Tele and Radio Research Institute, Centre of Advanced Technology, Ratuszowa 11, 03 -450 Warszawa, Poland 1

2

Abstract Purpose – The purpose of this paper is to characterise electrical properties of Ni-P thin-film resistors made on FR-4 laminate in a wide range of temperature (fro m -180 to 20ºC). Design/methodolog y/approach – The study was performed using resistors made of nickel–phosphorus (Ni-P) foil with two different thicknesses (0.1 μm or 0.05 μm) and the same different sheet resistances (100 oh m/sq or 250 ohm/sq). The resistance rectangular resistors had length and width fro m the range between 0.59 and 5.91 mm. The resistance vs temperature characteristics and their distribution as well as resistors durability to low-temperature thermal shocks were investigated.

Findings – The results showed almost linear temperature dependence of resistance with a negative TCR of about -95 ppm/°C for 250 ohm/sq layer and -55 ppm/°C for 100 ohm/sq. Very small dimensional effect was observed both for sheet resistance as well as for R(T) characteristic. Thin-film resistors are also characterized by very high durability to low-temperature thermal shocks. Research limitations/implications – These studies do not include more detailed determination of nature of observed characteristics and changes. There has not also been established what formulas are physically acceptable for description of experimental results. Originality/value – The results presented in this paper can be very useful for low-temperature applications of thin-film resistors made on PCB. They suggest possibility of wide applications of these components in a wide temperature range. Keywords: Thin-film resistor, Printed Circuit Board, Cryogenic temperature, Nickel-phosphorus (Ni-P) alloy Paper type: Research paper

1.

INTRODUCTION

Nowadays the miniaturisation of the electronic components, devices and circuits is still a driving force of technology. Since the classical solutions already reached their limits in term of the size reduction, other technologies are currently intensively developed. The future of increasing of electronics package density lies now in the domain of integration of components with the substrate. Mult ilayer technologies, like MCM-C, MCM-D and MCM-L, are generally used for these purposes [1-3]. The most widely developed are technologies concerning passive components, which are generally account for the largest number of elements in electronic circuits. The authors of this paper already conducted many investigations concerned with reliability, quality and stability of passives fabricated onto surface or embedded into printed circuit boards (PCBs) [4-6]. However these investigations were related with typical operating environments (mainly from room temperature to about 150°C). But an increasing interest in the field of low temperature electronics has been observed in recent years. On the other hand most manufacturers’ low temperature specifications end between -40 and -60°C [7]. Therefore we decided for broadening the temperature range of electrical characterization of Ni-P thin film resistors made onto PCB down to -180°C. Resistance, sheet resistance, their distribution and dependence on resistor aspect ratio as well as resistance versus temperature in the range from -180°C to 20°C, cold temperature coefficient of resistance (CTCR) and durability of resistors to thermal shocks between liquid nitrogen and room temperature are reported and analyzed in this paper.

P. Winiarski, A. Kłossowicz, W. Stęplewski, J. Wróblewski, A. Dziedzic; Electrical properties of thin-film resistors in a wide temperature range, Circuit World, vol.41(2015), p.116-120

2.

TEST STRUCTURES

The research subjects were thin-film resistors fabricated on FR-4 laminate with the aid of Ohmega-Ply® technology [6,8]. In this technique firstly thin layer of nickel-phosphorous (Ni-P) alloy is electroplated on copper foil. Next such composite foil, called RCM (resistor-conductor material), is laminated to FR-4 substrate. Finally copper circuitry and planar resistors are realized by subtractive (chemical etching) processes. In addition the surface of test structures (excluding contact pad areas) was covered by protective cladding. Two types of resistive foil - with sheet resistance 100 Ω/sq or 250 Ω/sq (with thickness 0.1 μm or 0.05 μm, respectively) - were used for fabrication of test structures. Single test structure consisted of 36 Ni-P rectangular resistors made on 50 30 mm2 FR-4 laminate. All pair combinations from the following set of values: 0.59, 1.31, 2.54, 3.96, 5.19, 5.91 mm were used as length and width of designed and fabricated resistors – the resistor aspect ratio n = l/w was changed in a very wide range between 0.1 and 10 (Fig. 1).

Fig. 1.

3.

Topology of test structures.

MEASUREMENT PROCEDURE

The resistance versus temperature measurement was performed in KSE-95 cryosystem [9] in a wide temperature range between -180°C and +20°C. To measure R(T) characteristics the resistor arrays were placed in a special probe holder with beryllium-copper gold plated pins mounted inside the vacuum insulated cryochamber supplied with nitrogen from the Dewar vessel and connected with measuring instruments by appropriate cables and connectors. The temperature was controlled by the flow of mass of liquid nitrogen which value was monitored by Pt100 sensor. All measurements were performed under NI Labview software and GPIB interface with as high accuracy as possible. To avoid parasitic effects short- and open-circuit corrections were made at the beginning of measurements. The resistance was measured at 100 kHz and 1 V amplitude, using HP4263A LCR meter with 5½ digits accuracy. To minimize the measurement noises the result of measurement is calculated as average value of 64 measurements taken at one point. The resistors were switched by Keithley 7001 scanner. The resistance versus temperature was measured in temperature range from -180°C to 20°C in steps of 10°C. Resistors were also subjected to series of hundred low-temperature thermal shocks (between room temperature and -196°C) with 10 min soak at every temperature (measurements were made after every 10 shocks). These investigations were carried out for resistors fabricated on FR-4 laminate (thin-film ones made of 100 Ω/sq or 250 Ω/sq nickel-phosphorus foil, polymer thick-film ones made of 200 Ω/sq ink from Electra D’Or company with Au or Cu terminations) and for cermet thick-film resistors fired on alumina at 850o C (made of 10 kΩ/sq R324 ink from Institute of Electronic Materials Technology, Warsaw).

4.

RESULTS AND DISCUSSION

4.1. Dimensional effect Ten resistor test coupons, each for them consisted of 36 resistors, were measured for every type of resistive foil. Before more complex analysis of results some basic statistical calculations were performed. The R(T) characteristics for each of 36 resistors were averaged over all 10 substrates. Also the standard deviations of R(T) characteristics were determined for every resistor at every measuring point. The dependence of sheet resistance (at 0°C) on resistor aspect ratio (n = l/w, where l – resistor length and w – resistor width) is shown in Fig. 2. Higher sheet resistance is observed for shorter and wider components. The largest differences between experimental and nominal values of sheet resistances in the investigated size range are equal to about 10% of designed value. The values of resistance were characterized by relatively small deviations

P. Winiarski, A. Kłossowicz, W. Stęplewski, J. Wróblewski, A. Dziedzic; Electrical properties of thin-film resistors in a wide temperature range, Circuit World, vol.41(2015), p.116-120 (typical values of variability coefficient vR are equal to about 1.0-1.5% - Tab. 1); in the worst cases they exceed 6% what occurred for some of the resistors with higher aspect ratio of dimensions. Tab. 1. Variability coefficient of resistance (vR[%] = sR/Rav 100%, where sR – standard deviation and Rav – mean value of resistance) at temperature 0°C (resistor dimensions in mm) 100 Ω/sq width length 5.91 5.19 3.96 2.54 1.31 0.59

250 Ω/sq

5.91

5.19

3.96

2.54

1.31

0.59

5.91

5.19

3.96

2.54

1.31

0.59

0.92 0.91 1.18 1.18 2.00 5.01

1.71 1.30 1.52 1.66 1.81 4.02

0.97 1.15 1.75 0.91 1.50 3.97

3.90 0.81 0.61 1.07 2.25 2.05

5.02 3.13 1.15 1.56 1.06 2.12

4.74 2.50 3.58 1.53 0.76 1.66

1.68 1.27 1.74 1.40 2.52 1.94

2.10 1.59 1.07 1.06 2.53 2.61

1.92 1.54 1.50 1.48 2.48 3.75

8.48 1.62 1.28 1.59 2.16 1.28

1.46 1.08 2.39 1.94 2.21 2.04

6.31 6.48 2.48 5.40 2.03 1.62

a)

b) 270

115 110

260

105

250 width [mm] 5.91 5.19 3.96 2.54 1.31 0.59

95 90 85 80

R/sq [

R/sq [

100

0

1

2

width [mm] 5.91 5.19 3.96 2.54 1.31 0.59

240

230

3

4

5

220

6

0

1

2

3

length [mm]

Fig. 2.

4

5

6

length [mm]

Sheet resistance as a function of resistor geometry for: a) 100 Ω/sq, b) 250/sq Ni-P resistive foil.

4.2. Temperature dependence of resistance All tested resistors exhibit almost linear decrease of resistance versus temperature. The examples of normalized resistance characteristics are shown in Fig. 3. The maximum changes in resistance (in the investigated range) relative to the temperature of 0°C were approximately 1% or did not exceed 2% for elements made of 100 Ω/sq or 250 Ω/sq Ni-P foils, respectively. a) b) 1.020

1.020

2

length×width (mm ) 5.91×5.91 5.91×5.19 5.91×3.96 5.91×2.54 5.91×1.31 5.91×0.59

2

length×width (mm ) 5.91×5.91 5.91×5.19 5.91×3.96 5.91×2.54 5.91×1.31 5.91×0.59

1.010

1.005

1.015

R(T)/R(0 C)

R(T)/R(0 C)

1.015

1.000

1.005

1.000

0.995 -200

-160

-120

-80

-40

Temperature [ C]

Fig. 3.

1.010

0

40

0.995 -200

-160

-120

-80

-40

0

40

Temperature [ C]

Example characteristics of normalized resistance versus temperature for a) 100 Ω/sq and b) 250 Ω/sq resistors.

P. Winiarski, A. Kłossowicz, W. Stęplewski, J. Wróblewski, A. Dziedzic; Electrical properties of thin-film resistors in a wide temperature range, Circuit World, vol.41(2015), p.116-120 Such behavior suggest that cold temperature coefficient of resistance (CTCR), calculated using the following formula CTCR = 106 ∙ [R(20°C) − R(−60°C)]/ [R(20°C)(20°C − (−60°C)], [ppm/ o C]

dR

is almost identical with differential TCR ( TCR diff

( R dT )

(1)

). This is quite different situation in comparison

with thick-film resistors. The values of CTCR and their dependence on resistor foil, size and aspect ratio are given in Tab. 2. This coefficient is equal to about -45 to -75 ppm/°C for resistors made of 100 Ω/sq foil and from -80 to -110 ppm/°C for resistors made of 250 Ω/sq foil, respectively. The influence of the components topology turns out to have rather insignificant impact on those factors. Tab. 2. Cold TCR (-60°C to 20°C) (resistor dimensions in mm) 100 Ω/sq width length 5.91 5.19 3.96 2.54 1.31 0.59

250 Ω/sq

5.91

5.19

3.96

2.54

1.31

0.59

5.91

5.19

3.96

2.54

1.31

0.59

-46 -56 -51 -50 -50 -47

-64 -56 -57 -60 -54 -70

-56 -56 -55 -53 -52 -45

-61 -53 -51 -49 -56 -75

-59 -52 -53 -49 -56 -61

-56 -60 -62 -43 -55 -47

-90 -100 -92 -89 -86 -81

-97 -96 -94 -92 -97 -107

-95 -98 -93 -88 -87 -85

-97 -97 -90 -88 -89 -87

-91 -89 -88 -90 -109 -102

-80 -81 -84 -86 -93 -86

4.3. Durability to low temperature thermal shocks The durability of tested structures were determined by relative changes in resistance R

R0

( R(n) R0 )

R0

(2)

100%

where: R0 – initial resistance (before shocks), R(n) – resistance after n shocks. The example results are presented in Fig. 4. a) b) 0,06

0,18 0,16

0,04 0,14 0,12

0,02

2

Dimensions [mm ] 1,31x5,91 1,31x1,31 1,31x0,59

R/R0 [%]

R/R0 [%]

0,10 0,00

-0,02

-0,04

Dimensions [mm ] 1,31x5,91 1,31x1,31 1,31x0,59

0,04

0,00 -0,02

-0,08

-0,04 0

20

40

60

80

100

0

Number of shocks

c)

0,06

0,02

2

-0,06

0,08

20

40

60

Number of shocks

d)

80

100

P. Winiarski, A. Kłossowicz, W. Stęplewski, J. Wróblewski, A. Dziedzic; Electrical properties of thin-film resistors in a wide temperature range, Circuit World, vol.41(2015), p.116-120 0,25

2,6 2

2

0,20

Dimensions [mm ] 1,31x5,91 1,31x1,31 1,31x0,59

Dimensions [mm ] 4x1 Au 1x1 Au 4x1 Cu 1x1 Cu

2,4 2,2 2,0

0,15

1,8 1,6

R/R0 [%]

R/Ro [%]

0,10

0,05

0,00

1,4 1,2 1,0 0,8 0,6

-0,05

0,4 0,2

-0,10

0,0 0

20

40

60

Number of shocks

Fig. 4.

80

100

0

20

40

60

80

100

Number of shocks

Relative resistance change during thermal cycling in the range -196°C to room temperature for: a) 100 Ω/sq Ni-P thin-film resistors, b) 250 Ω/sq resistors Ni-P thin-film resistors, c) R324 (10 kΩ/sq) cermet thick-film resistors, d) ED7500 (200 Ω/sq) polymer thick-film resistors.

Generally in literature there is very few information about such tests. Of course such behaviour is important for cryogenic resistance thermometers – please see e.g. [10]. Moreover durability of carbon/polyesterimide thick-film resistors to low-temperature thermal shocks was presented in [11]. Generally all tested thin-film resistors exhibit very high durability to low temperature thermal shocks – changes of resistance after 100 cycles were equal to about ±0.05% or +(0.08-0.15)% for resistors made of 100 Ω/sq or 250 Ω/sq foils, respectively. Mentioned fractional resistance changes are somewhat smaller than for thick-film cermet resistors (ΔR/R0 is changed between -0.10% and +0.20%) and decisively smaller in comparison with polymer thick-film resistors (ΔR/R0 of about 1.0% - 2.0% for structures with copper (Cu) contacts or 0.5% - 0.8% for copper contacts with Ni/Au protective coating).

5.

CONCLUSIONS

The dependency of electrical properties on temperature of thin-film resistors made of NiP foil was investigated in temperature range down to -180°C. The results showed linear changes of resistance determined by negative TCR. It turned out that resistors made of two times thinner film exhibits nearly two times stronger dependence of resistance on temperature. The results are consistent with other investigation of NiP resistors [12], adding that the planar dimension of the component has rather insignificant impact of its electrical properties at low temperatures. Thin-film resistors are also characterized by very high durability to low-temperature thermal shocks.

ACKNOWLEDGMENTS This work was supported by the National Science Centre (Poland), Grant DEC-2011/01/B/ST7/06564 and statutory activity of Wroclaw University of Technology.

6.

REFERENCES

1. Ulrich R.K., Schaper L.W., Integrated Passive Component Technology, Wiley Interscience – IEEE Press, 2003. 2. Jackson M., Pecht M., Lee Soon Book, Sandborn P., Integral, Embedded, and Buried Passive Technology, CALCE, University of Maryland (USA), 2003. 3. Dziedzic A., Electrical and structural investigations in reliability characterisation of modern passives and passive integrated components, Microelectronics Reliability, vol.42 (2002), pp.709–719. 4. Winiarski P., Dziedzic A., Kłossowicz A., Stęplewski W., Kozioł G., Analysis of long-term stability of thin-film and polymer thick-film resistors embedded in Printed Circuit Boards, Elektronika, vol.53 (2012), no.1, pp.55-58.

P. Winiarski, A. Kłossowicz, W. Stęplewski, J. Wróblewski, A. Dziedzic; Electrical properties of thin-film resistors in a wide temperature range, Circuit World, vol.41(2015), p.116-120 5. Winiarski P., Kłossowicz A., Stęplewski W., Nowak D., Dziedzic A., Analysis of steady-state and transient thermal properties of cermet, polymer and LTCC thick -film resistors, Circuit World, vol.40 (2014), pp.17-22. 6. Dziedzic A., Kłossowicz A., Winiarski P., Nitsch K., Piasecki T., Kozioł G., Stęplewski W., Wybrane właściwości elektryczne i stabilność elementów biernych wbudowanych w płytki obwodów drukowanych , Przegląd Elektrotechniczny, vol.87 (2011), no.10, pp.39-44. 7. Buchanan E.D., Benford D.J., Forgione J.B., Moseley S.H., Wollack E.J., Cryogenic applications of commercial electronic components, Cryogenics, vol.52 (2012), pp.550-556. 8. www.ohmega.com (Ohmega Technologies Inc. website). 9. Balik F., Sommer W., Environment for automated low-temperature measurements of electronic circuits, Elektronika, vol.52 (2011), no.3, p.84-89. 10. Żak D., Dziedzic A., Kolek A., Stadler A.W., Mleczko K., Szałański P., Zawiślak Z., Implementation of RuO2-glass based thick film resistors in cryogenic thermometry, Measurement Science and Technology, vol.17 (2006), pp.22-26. 11. Dziedzic A., Czarczyńska H., Licznerski B.W., Rangelow I.W., Further examinations of carbon/polyesterimide thick-film resistors, J. Materials Science: Materials in Electronics, vol.4 (1993), pp.233-240. 12. Józenków T., Dziedzic A., Borecki J., Kozioł G., Electrical and stability properties of thin-film resistors embedded in printed circuit boards, Elektronika, vol. 48 (2007), no.12, pp.23-26.