The influence of SO2 environments on immersion silver finished PCBs by mixed flow gas testing

2009 International Conference on Electronic Packaging Technology & High Density Packaging (ICEPT-HDP) 978-1-4244-4659-9/09/$25.00 ©2009 IEEE

The Influence of SO2 Environments on Immersion Silver Finished PCBs by Mixed Flow Gas Testing

Shunong Zhang1, Anshul Shrivastava2, Michael Osterman2, Michael Pecht2, 3, Rui Kang1 1Dept. of System Engineering of Engineering Technology

Beijing University of Aeronautics and Astronautics Beijing, 100191, P. R. China

2CALCE Center for Advanced Life Cycle Engineering University of Maryland

College Park, MD-20742, U.S.A 3Dept. of Electronic Engineering City University of Hong Kong

Abstract

This study focuses on the corrosion of immersion silver (ImAg) finished copper land patterns on printed circuit boards (PCBs) due to SO2 exposure in a mixed flow gas chamber. Six test conditions were examined with varying concentrations, temperatures, relative humidity, and exposure times. The results indicated that there are two mechanisms of corrosion on ImAg-finished PCBs in an SO2 gas environment: direct chemical corrosion and electrode reaction. No evidence shows that Ag2S and Cu2S or CuS were produced. In high humidity, chemical and electrode reaction both existed, and the corrosion products could included Ag2O, AgCl, Ag2SO3, CuO, CuCl2 and CuCl, In low humidity, the chemical corrosion was predominant, and the corrosion products could include Ag2O, CuO. Passive films were formed on ImAg finished surface under long exposure time. The temperature from 30°C to 40°C did not have an obvious influence on the ImAg-finished PCBs.

1. Introduction With the advent of the Restriction of Hazardous

Substances (RoHS) legislation in Europe, which forbids the use of lead in electronics products, the electronic industry has moved to lead-free surface finishes for printed circuit boards (PCBs). The widely used SnPb hot air solder level (HASL) process, has been replaced with organic solderability preservative (OSP), immersion silver (ImAg), electroless nickel immersion gold (ENIG), and immersion tin (ImSn) processes. Lead-free HASL is beginning to see an increased use. [1-7]

Among the HASL-free finishes, ImAg and OSP are the preferred finishes for many applications, while ImSn and ENIG are used for niche applications [4]. Due to inherent processing difficulties with OSP boards [3], and also due to the many desirable characteristics of ImAg such as remaining solderable for up to 12 months prior to assembly, having little effect on signal loss due to the good conductivity and thinness of silver, and manufacturing costs of plating the ImAg is half the price in comparison to ENIG and comparable with ImSn finishes [6-7], ImAg boards are quickly becoming a popular PC board finish in the electronics industry. However, ImAg has been shown to have issues in high sulfur environments [3].

Researchers have conducted studies on the corrosion phenomena and mechanisms of ImAg PCBs [1-6]. Some studies have also compared ImAg-finished PCBs with other

kinds of lead-free PCBs in the same batch experiments [1-2, 4-5]. Some researchers have used MFG tests to drive corrosion to conduct studies for the corrosion mechanisms of lead-free PCBs or evaluate its corrosion resistance.

On Cullen’s study [1], Creep corrosion was reproduced by using a condensing vapor test, which used H2S and added 1% hydrochloric acid to 0.1g/l sodium bisulfide to form a “sulfur chamber.” All finishes exhibited creeping corrosion within 24 hours under this environment. Cullen also conducted Class III MFG testing and found that creep corrosion did not occur until the humidity increased to over 93% to create condensation.

Veale [2] used the MFG test (100ppb H2S; 200ppb NO2; 200ppb SO2; 20ppb Cl2; temperature=28-29°C; 75%RH; test duration=20 days or 480 hours) to study the corrosion resistance of OSP, ImAg, ENIG, and ImSn. The progress of the test was monitored by copper reactivity, and the accumulated corrosion per day was 3500Angstroms/day, which is equivalent an ISA G1 (or Battelle class III). The results were that none of these coatings could be considered immune from failure in a Battelle class III environment, and ImSn and OSP could be expected to survive in a Battelle class II environment [4].

Xu et al. [4] conducted MFG testing on PCBs with OSP, ImAg, ImSn and ENIG finishes under a more severe environment (40°C; 69%RH; 1700ppb of H2S; 200ppb of NO2; 20 ppb of Cl2; and 200ppb SO2), which represented Battelle class IV and ISA class G2 conditions. The results after two days showed that all of the ImAg finishes were covered with grayish corrosion products, mostly Cu2S. Four out of seven of the ImAg finishes showed fiber-assisted electrochemical migration after five days of MFG exposure. One sample was observed to have creep corrosion along the fiber (fiber-assisted creep corrosion) after 10 days. From the test results, it was found that the ImAg-finished boards were more susceptible to corrosion than the other three types of boards. A second issue associated with ImAg is the blistering or peeling of the conductive corrosion products from the surfaces. After 10 days of the MFG test most of the samples only showed minor blistering. Peeling and flaking were only observed after 40 days of MFG exposure.

In no previous studies has a test been conducted in a corrosive environment containing a single corrosive gas, such as H2S or SO2, during the analysis and qualification of ImAg as a PCB finish. Our study in reference [8] focuses on the corrosion of ImAg finished copper land patterns on PCBs

116

2009 International Conference on Electronic Packaging Technology & High Density Packaging (ICEPT-HDP)

under H2S exposure. Twelve test conditions were examined with varying levels of H2S, temperature, relative humidity and exposure times. The results indicated both direct chemical reaction corrosion and electrode reaction corrosion, especially galvanic corrosion. Temperature shows significant influences on ImAg finished surface PCBs. Tests found extensive corrosion on ImAg finished PCBs at 40oC even in very low humidity. On ImAg finish surfaces, the corrosion is non uniform for the early period exposure, and corrosion modes mainly show as pitting, open mouth, particles, pits, the corrosion products mainly include Ag2O and Cu2O, as time goes on, the corrosion products mainly include CuS, Cu2S, CuO, Ag2S and form a passive film on the surface. ImAg finished PCBs are vulnerable to H2S gas, non uniform severe corrosion was found to occur at defects on the ImAg surfaces. MFG test can produce creep corrosion on ImAg finished PCBs by only using H2S gas. Dendrite corrosion products growing from the edge with solder mask usually is longer than from the edge without solder mask. The corrosion products of dendrites mainly included CuS or Cu2S.

The current study focuses on the influence of SO2 on ImAg-finished PCBs in MFG testing. Six case studies are presented, and the test results are compared.

2. Process of Samples Preparation For this study, segments of unassembled boards with

ImAg finish were used for test samples. The thickness of the ImAg finish is 0.23~0.36µm detected at 5 different locations by X-Ray Fluorescence Spectroscopy. An ImAg-finished PCB were segmented into many pieces using a Dremel tool, then were rinsed under running tap water, washed in an ultrasonic cleaner with de-ionized water, dried with ambient air and additional two hours in a temperature chamber (110°C). Before MFG testing, three pieces were selected for inspection using environmental scanning electronic microscope (ESEM Model: QUANTA 200) and Energy Dispersive X-ray Spectroscopy (EDS). Each PCB piece was approximately 50mm×50mm with through holes and pads. Bare copper samples were also prepared in order to monitor the corrosion rate. The copper samples were prepared as per section 7.6.1 of ASTM B810-01a [11].

Fig. 1(a) shows the typical ImAg finished PCB segment sample before tests. Fig. 1(b) shows a typical ImAg surface under high magnification.

(a) (b)

Fig. 1 The typical ImAg finished surface before tests The elements on ImAg surface include Ag, Cu by a point

EDS analysis under 30kv, and the elements on solder mask include C, O, Si, S, Ba, Ca. These findings were reconfirmed

in a separate SEM(S-530) & EDS (Link ISIS) system under high vacuum and 10kV, Fig. 2 (a) and Fig. 3 show a point analysis on ImAg finished surface, and Fig. 2 (b) and Fig. 4 show a point analysis on solder mask surface.

(a) (b)

Fig. 2 Points analysis on ImAg finished surface and solder mask surface before tests

Fig. 3 A point analysis on ImAg finished surface by EDS

(10kV)

Fig. 4 A point analysis on solder mask surface by EDS (10kV)

3. Experiments and Results Six test conditions, outlined in Table 1, were examined in

this study. For each test, the PCB pieces and copper coupons were suspended by insulated wires in the MFG chamber. During the exposure, visual observations were conducted. ImAg samples were taken out on the exposure times listed in Table 1. Upon completing the set exposure time, test specimens were examined by ESEM (QUANTA 200) and their compositions were detected by EDS in order to study their characteristics. Some samples were also examined by ESEM (Model: QUANTA 600) & EDS, and by SEM (S-530) & EDS (link ISIS) again when needed.

117

2009 International Conference on Electronic Packaging Technology & High Density Packaging (ICEPT-HDP) 978-1-4244-4659-9/09/$25.00 ©2009 IEEE

Table 1 Test Matrix ① ② ③ ④ ⑤ ⑥ ⑦

1 30±1 75±1 70±30 5, 24, 48, 72, 120, 144, 168, 192, 216, 240

30 8

2 32±1 >95 50±20 72, 120, 168 15 5 3 32±1 <5 200, 400 48, 96 8 2 4 40±1 <5 400 48, 72 6 2 5 40±1 30 100±10 48 4 2 6 30±1 75±1 500±100 4.5 1 1

①Test; ②T (°C); ③%RH; ④Gas concentration (ppb); ⑤Exposure time (hours); ⑥Number of ImAg samples; ⑦Number of Cu samples.

The SO2 concentrations of Test 1~Test 5 in Table 1 were

obtained by calculation according to flow rate of SO2 and air, because the instrument calibrating the concentration of SO2 was failure at that time. The SO2 concentration of Test 6 was obtained by new equipment: the Honeywell Single Point Monitor. The theory of calculating the SO2 concentration in an MFG chamber is shown in Fig. 5, and the formula is shown in Equation (1).

Fig. 5 The theory of calculating SO2 concentration in MFG

chamber

2

2

22

[ ][ ] STD SO

chamberD SO

SO FSO

F F•

=+

(1)

2[ ]chamberSO : the estimated SO2 concentration in the MFG chamber, ppm;

2[ ]STDSO : concentration of the cylinder SO2 standard, ppm, here was 1196 ppm;

2SOF : flow rate of the SO2, slpm;

DF : flow rate of the air, which include dry air and wet air, slpm.

Fig. 6 shows the samples of last exposure time in test 1~test 6. The samples did not show obvious change but looked a little bit faded in long exposure time (Fig. 6(a)~(e))

(a)240 hours in

test 1 (b)168 hours in

test 2 (c) 96 hours in

test 3

(d) 72 hours in

test 4 (e) 48 hours in

test 5 (f) 4.5 hours in

test 6 Fig. 6 Samples of exposure in test 1 ~ test 6

(1)Test 1 The first test condition included a temperature of 30±1°C,

a relative humidity of 75±1%RH, and an estimated concentration of SO2 of 70±30ppb. The exposure duration was 10 days. After 5, 24, 48, 72, 120, 144, 168, 192, 216 and 240 hours, at least one sample was taken out to check on these exposure times. The surfaces of samples after 240 hours exhibited many dark and bright points under ESEM (QUANTA 200). A point analysis by EDS on a dark point showed that the composition included Ag, Cu, and Cl (Fig. 7 (a) and Fig. 8). Another point analysis on a bright point showed that the composition included Ag, Cu, Cl, O, and C (Fig. 7 (b) and Fig. 9).

(a) (b)

Fig. 7 The surface of sample after 10 days exposure time in test1

118


Fig. 8 A point analysis on a dark point after 10 days in test 1 (30kV)

Fig. 9 A point analysis on a bright point after 10 days in test 1

(30kV) (2)Test 2 The second test condition included a temperature of 32 ±

1°C, a relative humidity of > 95%RH, and an estimated concentration of SO2 of 50 ± 20ppb. The exposure duration was 7 days. One sample was removed for inspection at 72, 120, and 168 hours. The samples after 168 hours also exhibited many dark and bright points, and corrosion products were found at the edges by ESEM (QUANTA 200) (Fig. 10 (a)). A 120 hours sample was checked by ESEM&EDS(Fig. 10 (b)), a point analysis by EDS on a bright point showed that the composition included Ag, Cu, Cl, S,O,C (Fig. 11).

(a)168 hours (b) 120 hours

Fig. 10 168 hours and 120 hours ImAg surface in test 2

Fig. 11 A point analysis on a bright ‘cotton’ point after 120 hours in test 2 (30kV)

(3)Test 3 The third test condition included a temperature of 32 ±

1°C, a relative humidity of < 5%RH, and an estimated concentration of SO2 of 200-400ppb(at first the concentration is 200ppb, after 48 hours it was changed to 400 ppb). The exposure duration was 4 days. One sample was checked after 48 and 96 hours. The surface of sample after 48 hours of exposure in test 3 showed almost no changes under ESEM (Fig. 12(a)), and the surface of the sample after 96 hours of exposure time shows a small number of dark points (Fig. 12(b)). A point analysis by EDS on a dark region showed that the composition included Ag, Cu, O, and C (Fig. 13).

(a) 48 hours (b) 96 hours Fig. 12 The surfaces of samples after 48 and 96 hours

exposure time in Test 3

Fig. 13 A point analysis on a dark region after 96 hours in test

3 (30kV)

(a) 48 hours (b) 72 hours

Fig. 14 The ImAg surfaces after 48 and 72 hours of exposure time in test 4

(4)Test 4 The fourth test condition included a temperature of

40±1°C, a relative humidity of < 5%RH, and an estimated concentration of SO2 of 400ppb, the exposure duration was 3 days. One sample was removed for inspection after 48 and 72 hours. Checking these samples with ESEM (QUANTA 200), the surface of the samples after 48 hours of exposure time showed some dark points (Fig. 14 (a)). The surface of the

119


sample after 72 hours of exposure showed more dark points (Fig. 14 (b)).

(5) Test 5 The fifth test condition included a temperature of 40±1°C,

a relative humidity of 30%RH, and an estimated SO2 concentration of 100ppb, the exposure duration was 2 days.

Checking the sample under ESEM (QUANTA 600) & EDS under high vacuum status, Sulfur was difficult to found on the surface (e.g. Fig. 15 (a) and Fig.16). A point analysis on the surface by SEM(S-530) & EDS (link ISIS) also showed that the surface was clear, and the composition only included Ag and Cu (Fig. 15 (b) and Fig. 17).

(a) (b)

Fig. 15 ImAg surface under ESEM (QUANTA 600) and SEM (S-530) in test 5

Fig. 16 A point analysis under high vacuum status by ESEM

(Model: QUANTA 600) and EDS in test 5 (15kV)

Fig. 17 A point analysis under high vacuum status by SEM

(S-530) and EDS (link ISIS) in test 5 (20kV) (6) Test6 The sixth test condition included 30±1°C, 75%RH and

500±100ppb SO2. The exposure duration was 4.5 hours. A point analysis on the surface edge by ESEM (QUANTA 200) & EDS showed that the composition included Ag, Cu, Si, S, O, C, and Ba (Fig. 18(a) and Fig. 19), Si, C, Ba and some of

O, S should come from solder mask (Fig. 4). However, S was difficult to found on the ImAg surface by SEM (S-530) & EDS (link ISIS) (e.g. Fig. 18 (b) and Fig. 20).

(a) (b)

Fig. 18 ImAg surface under ESEM (QUANTA 200) and SEM (S-530) in test 6

Fig. 19 A point analysis by ESEM (QUANTA 200) & EDS in

test 6 (15kV)

Fig. 20 A point analysis by SEM(S-530) & EDS (link ISIS) in

test 6 (20kV)

4. Discussion In these case studies, Chlorine was found in the corrosion

products (test 1 and test 2). The Cl2 cylinder was not connected the tube going to the flow meter when these experiments were conducted. Where did the Chlorine come from? Judging from the gas and air path, it should come from the water that was in the heater barrel for the purpose of increasing the humidity. However, the water was distilled water, which came from a distilled water instrument. But anyway, Chlorine should not be much more if it came from the water. So, there are SO2, NO2, O2 and a small amount of Cl2 in the MFG chamber. SO2 is easy to dissolve in water, in the presence of high humidity, equations below exist [21]:

SO2 (g) = SO2 (aq) (2) H2O +SO2 (aq) = H2SO3 (aq) (3)

H2SO3 (aq) = H+ + HSO3- (4)

120


HSO3- = H+ + SO3

2- (5) Other species also were generated, the ImAg surface in the

MFG chamber constituted a rather complex electrode reactions systems, equations below exist [21-22]:

NO2 (g) = NO2 (aq) (6) 2NO2 (aq) + H2O = HNO2 (aq) + H+ + NO3

- (7) Cl2 (g) = Cl2 (aq) (8)

Cl2 (aq) + H2O = HOCl (aq) + H+ + Cl- (9) 3HOCl (aq) = 2Cl- +ClO3

- + 3H+ (10) 4H+ + NO3 +3Cl- = NOCl (aq) + Cl2 (aq) +2H2O (11)

Ag = Ag+ + e (12) O2 + H2O +4e = 4OH- (13)

2Ag+ + 2OH- = 2AgOH =Ag2O + H2O (14) Ag+ + SO3

2- = Ag2SO3 (s) (15) Ag+ + Cl- = AgCl (s) (16) Ag + Cl- = AgCl (s) + e (17)

Because of the porosity of the ImAg finish, the following equations could exist:

Cu = Cu+ + e (18) Cu = Cu2+ + 2e (19) Cu+ = Cu2+ + e (20)

2Cu + H2O = Cu2O(s) + 2H+ + 2e (21) Cu2O(s) + H2O = CuO(s) + 2H+ + 2e (22)

Cu2O(s) + H2O + 2OH- = 2Cu (OH)2(s) +2e (23) 2Cu (OH)2 (s)= CuO (s) + H2O(l) (24)

Cu2+ + 2Cl- = CuCl2 (s) (25) Cu + Cl- = CuCl (s) + e (26)

2CuCl2 (aq)+SO2 =2CuCl(s) + 2HCl(aq) + H2SO4(aq) (27) Some chemical reactions may also exist:

Ag + O2 =Ag2O (28) Cu +O2 = CuO (29)

Cu + O2 = Cu2O (30) Cu + CuCl2 = 2CuCl (31)

6CuCl + 3/2O2 + 3H2O = 2Cu3Cl2(OH)4 + CuCl2 (32) Ag + SO2 = Ag2S + O2 (33) Cu +SO2 = CuS +O2 (34)

However, Fig. 13 in test 3 shows that Equations (33) and (34) were not likely to occur, the corrosion products if existed were mainly Ag2O or CuO, Cu2O rather than Ag2S or CuS.

In high humidity, chemical and electrode reaction should have been present and the corrosion products mainly should have been Ag2O, CuO, Cu2O, Cu (OH)2, Ag2SO3. The corrosion products of AgCl, CuCl2, CuCl, etc. should also existed because Cl- existed in water, even its amount is lower.

AgCl and Ag2SO3 both are white in color [23-24]; Ag2O is a fine black or dark brown powder [25]; CuCl2 is a yellow-brown solid [29]; CuCl is a colorless solid [30]; and CuO are black-colored powder [26]; Cu2O is a brick red color[27]; Cu (OH)2 is a pale blue, gelatinous solid [28], and none of them dissolves easily in water. Fig. 6 did not show what kind of corrosion products are much more than others, but Cu2O and Cu (OH)2 should not be much.

Long exposure time did not show severe corrosion on the ImAg surface, it should because passive films of these corrosion products above were formed on corrosion points or corrosion surface, and they will prevent gas in MFG chamber from producing corrosion again.

5. Conclusions 1) There should be two mechanisms of the corrosion of

ImAg-finished PCBs in an SO2 gas environment: direct chemical reaction and electrode reaction.

2) No evidence shows that Ag2S and Cu2S or CuS existed. 3) In high humidity, chemical and electrode reaction both

existed, and the corrosion products could included Ag2O, AgCl, Ag2SO3, CuO, CuCl2, CuCl, etc.. In low humidity, the chemical corrosion was predominant, and the corrosion products could include Ag2O, CuO.

4) Passive films were formed on ImAg finished surface under long exposure time.

5) The temperature from 30°C to 40°C did not have an obvious influence on the ImAg-finished PCBs.

References 1. Cullen D., “Surface Tarnish and Creeping Corrosion on

Pb-free Circuit Board Surface Finishes,” IPC Works, 2005.

2. Veale, R., “Reliability of PCB alternate Surface Finishes in a Harsh Industrial Environment,” SMTA, 2005

3. Mazurkiewicz, P., “Accereated Corrosion of Printed Circuit Boards Due to High Levels of Reduced Sulfur Gasses in Industrial Environments,” Proceedings of the 32nd International Symposium for Testing and Failure Analysis Nov.12-16, Renaissance Austin Hotel, Austin, Texas, USA pp. 469-473, 2006.

4. Xu C., Flemming D.,Demerkin K., “Corrosion resistance of PCB Surface Finishes,” Alcatel -Lucent, Apex, 2007.

5. Schueller R., “Creep Corrosion on Lead-free Printed Circuit Boards in High Sulfur Environments,” pp. 643-654, 2007.

6. Zhou, Y.L., Pecht, M., “Assessment of Immersion Silver Finished Circuit Board Assemblies Using Clay Tests,” Proceedings of ICRMS2009: International Conference on Reliability Maintainability and Safety, Chengdu, China, July 21-25, 2009.

7. Wang, W.Q., Choubey,A., Azarian, M, Pecht, M., “An Assessment of Immersion Silver Surface Finish for Lead-free Electronics,” Journal of Electronic Materials, 2009.

8. Zhang, S.N., Osterman,M., Shrivastava, A., Pecht, M., Kang, R., “The Influence of H2S Environments on Immersion Silver Finished PCBs by Mixed Flow Gas Testing,” IEEE Transaction on Device and Material Reliability, submitted.

9. ASTM Designation: B845-97: Standard Guide for Mixed Flowing Gas (MFG) Tests for Electrical Contacts.

10. ASTM Designation: B827-05: Standard Practice for Conducting Mixed Flowing Gas (MFG) Environment Tests.

11. ASTM Designation: B810-01a: Standard Test Method for Calibration of Atmospheric Corrosion Test Chambers by Change in Mass of Copper Samples.

12. Abbott, W.H., “The Development and Performance Characteristics of Mixed Flowing Gas Environment,” IEEE Trans.Components, Hybrids, Manufacturing Technol., Vol. 11, No.1, March 1988, pp. 22-35.

121


13. ISA-S71.04-1985, “Environmental Conditions for Process Measurement and Control System: Airborne Contaminants,” Instrument Society of America, 1985.

14. Telcordia GR-63-CORE Issue 2, Section 5.5, “Airborne Contaminants Test Methods,” Nov. 2000.

15. CALCE Standard Operating Procedures, Mixed Flow Gas Chamber-Rev.A, September 1999.

16. CALCE Standard Operating Procedures for ESEM, [Model:Quanta 200F(26A6)].

17. Precision Calibration System Instruction Manual: Model 491 Interim. Kin-Tek Laboratories, Inc.

18. Operating Instructions: Trace Source ULED Permeation Sources. Kin-Tek Laboratories, Inc.

19. OptiSonde General Easten Chilled Mirror Hygrometer User’s Manual.

20. 8270_74_80_84 Mass-Flow Control Boxes Instruction Manual, Matheson Gas Products.

21. Olof A. Svedung, Lars-Gunnar Johansson, and Nils-Gosta Vannerberg, “The Influence of NO2 and Cl2 at Low Concentrations in Humid Atomospheres on the Corrosion of Gold-Coated Contact Material,” IEEE Transactions on Components, Hybrids, and Manufacturing Technology, Vol. CHMT-9, No.3, Sept. 1986.

22. Cao, C.N., Principles of Electrochemistry of Corrosion, Chemistry Industry press, Beijing, Feb. 2008.

23. Silver chloride, http://en.wikipedia.org/wiki/AgCl 24. silver sulfite,

http://www.chemyq.com/xz/xz7/60855ennbm.htm 25. Silver oxide, http://en.wikipedia.org/wiki/Ag2O 26. Copper(II) oxide, http://en.wikipedia.org/wiki/CuO 27. Copper(I) oxide, http://en.wikipedia.org/wiki/Cu2O 28. Copper(II) hydroxide,

http://en.wikipedia.org/wiki/Cu(OH)2 29. Copper(II) chloride, http://en.wikipedia.org/wiki/CuCl2 30. Copper(I) chloride, http://en.wikipedia.org/wiki/Cucl

122

The influence of SO2 environments on immersion silver finished PCBs by mixed flow gas testing

Documents