Hair color changes caused by dyeing and

j. Csmet. Sci., 55, 437447 (September/Octber 2004) Hair clr changes caused by dyeing and thermal treatments ANA CAROLINA SANTOS NOGUEIRA, CARLA SCANAVEZ, CRISTIANE CARNELOS, ALEXANDRE GASPARI, and INI S JOEKES, Institut de Qu/mica, Universidade Estadual de Campinas (UNICAMP), Caixa Pstal 6154, 13084-971 Campinas, SP, Brazil. Accepted fr publicatin August 26, 2004. Presented at the 22 " IFSCC Cngress, Edinburgh, Sctland, September 2002. Synpsis The aim f this study was t shw the effect f heat expsure, dyeing, and shamping n hair clr as measured by diffuse reflectance spectrphtmetry. Successive dyeing f virgin hair with six permanent cmmercial frmulatin shwed that clr saturatin was btained after the first dyeing cycle. An unexpectedly high difference in hair clr saturatin, measured as DE* values, was btained fr virgin hair samples that differed nly in cleansing histry. After six sequential washings f the dyed hair samples, n difference was bserved in clr durability, indicating that the adhesin strength is similar t lng-lasting and tne-up dyeing frmulatins. Expsure t a ht plate at 172øC shwed a significant darkening f the virgin hair samples after 2 min. On the ther hand, virgin hair samples expsed t the gentler heat f a hand dryer (-60øC) shwed partial disappearance f the hair medulla after 60 min. Hwever, values f ttal clr difference were near the errr limit. INTRODUCTION Hair health is frequently related t its clr and glss. Hair clr is attributed t the melanin granules present in the crtex (1). Hwever, nly a few papers are fund in the literature invlving hair clr. This paper describes the clr prperties f hair as related t sme simple csmetic treatments. McMullen and Jachwicz (2) studied the effect f curling irns n hair clr. Samples were submitted t a ht plate at 160øC fr 30 min. Clr analysis was dne using the Hunter L*, a*, b* clr crdinates. Significant clr changes were bserved, such as an increase in yellwness f white and Piedmnt hair and yellwing and darkening f bleached hair. The mst frequent glss measurement is the specular glss, which crrespnds t the Address all crrespndence t In s Jekes. 437

438 JOURNAL OF COSMETIC SCIENCE light speculareflectin f a sample as cmpared with a standard (3,4). The decrease in surface rughness has direct influence n glss. Sauremann et al. (5) bserved that the hair rughnesset by the cuticles can be increased with their deteriratin r reduced by shamping. Studies using transmissin electrn micrscpy shwed that cnsecutive shamping cycles extract small amunts f prteins frm the endcuticle, leaving empty spaces (hles)(6,7,9). Shamps can als disslve structuralipids and prteinaceus material frm the cell membrane cmplex (8). Other treatments such as brushing, cmbing, and twel drying als damage the cuticles, and can generate hles in the endcuticle, rupturing and detaching the external cuticles, until the ends split. We have shwn that these types f damage prmte changes in human hair lightness (9). MATERIALS AND METHODS HAIR SAMPLES Tresses f virgin dark-brwn hair, each weighing 2.0 g and apprximately 15 cm in length, were used thrughut. The hair samples were aligned frm rt t tip end and tied near the rt end. Sme f them were cleansed by 8-h extractin with ethyl ether in a Sxhlet, dried at rm temperature, cmbed, and stred in a black desiccatr prir t the treatments. DIFFUSE REFLECTANCE SPECTROPHOTOMETRY The diffuse reflectance measurements were perfrmed using a Macbeth Clr-Eye 2020 diffuse reflectance spectrphtmeter. The instrument peratin cnditins were: (a) cnfiguratin CRIIS (C: white ceramic calibratin, R: reflectance, I: ultravilet waves included, I: specular cmpnent included, S: shrt viewing aperture); (b) D65 illuminant; and (c) internal reference. The peratin and the measurement cnditins fr human hair were established in a previus wrk (10). Spectra prvided values f crdinates L* (clr lightness), a* (redness if psitive, r greenness if negative), and b* (yellwness if psitive, r blueness if negative) frm the CIELAB system f equatins. Frm these, the clr difference parameters, DL* (lightness difference), Da* (red-green difference), Db* (yellw-blue difference), and DE* (ttal clr difference) were calculated. Measurement errr in ttal clr difference was abut DE* = 0.1, and hair sample variability dispersin was abut DE* = 1.0. Measurements were dne, keeping the same sample regin and turning the hair sample in the instrument sample hlder. The internal reference is ne chsen measurement f this set that is clsest the average values fr the set. Ten diffuse reflectance measurements were dne in each tress, and the values f the clr parametershwn in the tables are arithmetic averages and estimated standard deviatins f these sets. LIGHT MICROSCOPY Ten hair fibers 6-cm lng were psitined side by side in glass slides. Rt and tip regins were bserved under white light in a bright field using a Leica MZ 12.5 stere

EFFECT OF DYEING AND HEAT ON HAIR COLOR 439 micrscpe. Observatins were registered in TMAX ISO 100 Kdak film, rendering a set f 80 phtmicrgraphs. PERMANENT DYEING Six cmmercial permanent dyeing frmulatins frm L'Oreal Paris were used, three lng-lasting frmulatins (Imedia Excellance Cr6me, cntaining ammnia and hydrgen perxide) and three tne-up frmulatins (Casting, ammnia free, and lw hydrgen perxide). Fr each lng-lasting frmulatin, a similar clr f tne-up frmulatin was chsen. The frmulatins were f black, blnd, and red clrs. Each dye was applied t tw tresses, fllwing the enclsed instructins. All prcedures were dne using glves and after wetting the hair. The lng-lasting frmulatin was left acting fr 40 min, and the tne-up frmulatin was left acting fr 20 min. After this time, the tresses were washed in warm water and dried at rm temperature. The virgin hair tresses were dyed three successive times t attain clr saturatin; DRS measurements were taken after each dyeing cycle. After the three dyeing cycles, each tress was submitted t six sequential shamping washes. A cmmercial L'Oreal shamp frmulatin particularly indicated fr dyed hair was used. T each wetted tress, 1 ml f shamp was applied with glves and rubbed in fr 5 min; the tress was washed ut in tap water and dried at rm temperature. DRS measurements were taken after each shamping cycle. EFFECT OF EXPOSURE TO A HOT PLATE A Taiff prfessinal ht plate was used. The wrking temperature was 172øC, as measured by a thermcuple. Tw virgin hair tresses were used. The ht plate was applied as fllws: each hair tress was expsed t the plate fr 15 s and then cled fr 10 s, fur times. This gives a 1-min expsure time. The tresses were then washed with distilled water fr 20 s and dried with a hand dryer fr apprximately 7 min. This treatment was repeated up t a 10-min expsure t the ht plate. DRS measurements were dne after each 1-min expsure. The silicne treatment was dne in tw tresses previusly washed with distilled water. T each tress, 2 ml f prduct was applied, being manually distributed frm the rt t the tip end f the tress. These tresses, dried at rm temperature, were submitted t the same ht-plate expsure treatment f the untreated hair tresses. The silicne used was Dw Crning 1501 fluid, a csmetic ingredient recmmended fr renewing hair split ends. The prduct data sheet specifies that it cntains cyclpentasilxane and dimethicnl. EXPOSURE TO DRYER HEAT A Brittania 1300W dryer was used. The average wrking temperature was 62øC, as measured by a thermcuple. A virgin hair tress was attached t a wd hlder using cttn yarns. The dryer was psitined in frnt f the tress, at middle height, keeping a distance f 6 cm between the tress and the dryer. The tress was expsed fr 10 min in five cycles, at 20-min intervals. After 12 h at rm temperature, the same hair tress was cntinuusly expsed t the dryer fr 60 min and DRS measurements were taken. Fibers were cllected frm the virgin hair tress befre (cntrl) and after the heat treatments fr light micrscpy bservatins.

440 JOURNAL OF COSMETIC SCIENCE (a) ½"5, ILl 4- [] A 02 1! I Dyeing times (b) [] [] [] A 0 A _ 0 ' I I I I 0 ; 2 3 4 6 Shamping times Figure 1. Average ttal clr difference (DE*) values measured in six virgin brwn hair tresses. (a): Hair tresses were dyed three sequential times with three lng-lasting and three tne-up dying frmulatins f black, red, and blnd clrs. (b): The dyed tresses were submitted t six sequential shamping cycles. Reference values: clr parameters measured in the same tress befre dyeing. Symbls: (I) lng-lasting black, ([ ) tne-up black, (A) lng-lasting red, (/ ) tne-up red, (0) lng-lasting blnd, and (O) tne-up blnd.

EFFECT OF DYEING AND HEAT ON HAIR COLOR 441 12 10, [] m m [] _ (a) _ & A 1 I I I Dyeing times 12 lo [] (b) /x A ½D 2- i!! i i i 0 1 2 3 4 5 6 Shamping times Figure 2. Average ttal clr difference parameter (DE*) values measured in six virgin brwn hair tresses previusly cleansed with ethyl ether. (a): Hair samples were dyed three times with three lng-lasting and three tne-up dying frmulatins f black, red, and blnd clrs. (b): The dyed tresses were submitted t six sequential shamping cycles. Reference values: clr parameters measured in the same tress befre dyeing. Symbls: (m) lng-lasting black, ([ ) tne-up black, (') lng-lasting red, (/ ) tne-up red, (0) lng-lasting blnd, and (O) tne-up blnd.

442 JOURNAL OF COSMETIC SCIENCE Table I Clr Difference Parameters f Brwn Hair Samples Expsed t Ht Plate at 172øC (average temperature) Sample 1 Sample 2 Expsure time (min) DL* Da* Db* DE* DL* Da* Db* DE* 1-2.56-0.75-1.01 2.86-1.41-0.27-0.35 1.48 2-4.71-0.82-1.10 4.91-4.25-0.34-0.26 4.28 3-5.26-0.58-0.67 5.34-4.57-0.29-0.07 4.57 4-5.08-0.71-0.68 5.18-4.89-0.20-0.17 4.90 5-5.25-0.39-0.60 5.30-4.21-0.20-0.37 4.23 6-4.82-0.33-0.38 4.86-3.69-0.38-0.41 3.75 7-5.18-0.79-1.20 5.38-5.20-0.32-0.51 5.24 8-5.50-0.68-1.16 5.66-4.23-0.31-0.44 4.28 9-5.48-0.38-0.64 5.54-4.14-0.09-0.38 4.16 10-5.37-0.50-0.90 5.47-4.80-0.35-0.56 4.85 Sample 1 rej rence values: L* = 29.72, a* = 2.26, b* = 2.35. Sample 2 re rence values: L* = 28.62, a* = 2.06, b* = 2.08. Table II Clr Difference Parameters f Brwn Hair Samples Treated With Silicne and Expsed t Ht Plate at 172øC (average temperature) Sample 1 Sample 2 Expsure time (min) DL* Da* Db* DE* DL* Da* Db* DE* 1 0.07-0.48-0.44 0.92 1.42-0.05-0.17 1.47 2 0.80-0.32 0.07 0.90 1.98-0.08-0.14 2.00 3 0.59-0.60-0.72 1.25 0.80-0.21-0.90 0.88 4 0.90-0.64-0.68 1.35 1.54-0.14-0.23 0.52 5 1.02-0.43-0.49 1.21 0.86 0.03-0.06 0.99 6 0.74-0.52-0.67 1.21 0.97 0.01 0.05 1.00 7 0.64-0.50-0.27 1.01 1.01-0.05 0.08 1.13 8 0.55-0.37-0.40 0.85 1.05 0.04 0.17 1.09 9 0.34-0.50-0.63 0.92 0.99 0.22 0.41 1.15 10 0.47-0.33-0.40 0.77 0.76-0.10 0.04 0.83 Sample 3 rej rence values: L* = 21.77, a* = 2.44, b* = 2.62. Sample 4 rej rence values: L* = 21.51, a* = 1.76, b* = 1.85. RESULTS AND DISCUSSION EFFECT OF DYEING AND SHAMPOOING CYCLES The dyeing experiments were dne t quantify clr saturatin in hair using cmmercial dyeing prducts and prcesses. Secndly, cnsecutive shamping cycles f dyed hair samples were perfrmed t bserve dye persistence in hair. Permanent dyeing using lng-lasting and tne-up frmulatins were applied t the virgin hair cleansed r nt cleansed with ethyl ether. Figures 1 and 2 shw the clr difference parameter (DE*) btained after dyeing and subsequent shamping f hair samples nt extracted with ether and f hair previusly

EFFECT OF DYEING AND HEAT ON HAIR COLOR 443 Table III Ttal Clr Difference Parameter f Brwn Hair Sample Expsed t Dryer in Steps f 10 Min, with 20-Min Intervals, and After 60 Min f Cntinuus Expsure Expsure time (min) Hair regin DE* Expsure time (min) Hair regin DE* 10 Rt 0.9 + 0.2 40 Rt 1.0 + 0.4 Middle 0.7 + 0.3 Middle 0.9 + 0.3 20 Rt 0.9 + 0.3 50 Rt 1.1 + 0.4 Middle 0.6 + 0.2 Middle 1.1 + 0.4 30 Rt 1.2 + 04 110# Rt 1.3 + 0.2 Middle 1.0 + 0.2 Reference values: Rt: L* = 21.3, a* = 3.2, b* = 3.2, Middle: L* = 1.8, a* = 3.6, b* = 3.4. # Rej rence values: Rt: L* = 21.0, a* = 3.5, b* = 3.7. extracted with ethyl ether. The black dye frmulatins rendered the higher clr change (DE* = 6.0 and 10.0) in the dark brwn hair samples, fllwed by frmulatins f red (DE* 4.0 and 7.0) and blnd (DE* 1.5 and 4.0) dyes, respectively, t the ntcleansed and the earlier ether-cleansed hair samples. This was expected, since n bleaching treatment was dne n the brwn hair befre dyeing. Hair clr saturatin was bserved after the first dyeing fr bth the tne-up and lng-lasting frmulatins. N significant changes in the DE* values were bserved in the dyed hair samples after the six shamping cycles. Despite differences in the dying time and frmulatins, the dyes srbed in the hair during the dying prcesses were nt washed ut. The DE* values shwn in Figure 2 are rughly twice thse in Figure 1, shwing that ethyl ether cleansing strngly acts upn the srptin f dyes in hair. This is unexpected, since cleansing with ethyl ether shuld have nly a small effect n the hair srptin prperties. Nte that n significant changes in DE* values were bserved in the dyed hair samplesubmitted t the six shamping cycles. Cmparisn between Figures 1 and 2 shws that the amunt f dye remved frm dyed hair sample seems t be similar despite the hair cleansing histry. Therefre the strength f the dyes' srptin frces is alike in bth ethyl-ether-cleansed and nt-cleansed hair samples. Permanent dyes cnsist f dye precursrs, dye cuplers, and an xidizing agent reacting at alkaline ph 8-10 (8). Early studies shwed that hair cleansing with ether causes an increase in the percentage f empty cavities in the hair endcuticle when cmpared t hair treated with surfactant slutin and als ht water (9). The ethyl ether cnsequence must be then t pen a way fr the diffusin f the reactive substances in hair fiber by the permanent dye frmulatins. EFFECT OF EXPOSURE TO A HOT PLATE The effect f a ht plate n hair clr is shwn in Tables I and II. The DE* values in Table I shw that a cnsiderable clr change takes place when the hair is expsed t a ht plate. This clr change is bserved as sn as a 2-min expsure, remaining cnstant after 3 min. It is imprtant t ntice that clr changes are particularly related t the DL* values, which, being negative, indicate that all tresses

.. 444 JOURNAL OF COSMETIC SCIENCE medulla.. - i A bqir fiber B 340 nm MF MF C Figure 3. Phtmicrgraphs f virgin dark brwn hair shwing the medulla in different views. A: Hair fiber seen under reflected light in the steremicrscpe at 660x. B: Hair fiber seen under transmitted light in the steremicrscpe at 660x. The arrw pints t the cntinuus medulla placed in the middle f the hair fiber. C: Transmissin electrn micrgraph f virgin dark-brwn hair fiber (Elmiskp101 at 80 kv, smium-stained) shwing the crtex regin. Nte the nn-fibrillate cnstitutin f the medulla (M), which is surrunded by the macrfibrills (MF).

. EFFECT OF DYEING AND HEAT ON HAIR COLOR 445 A c discntinue medulla Fig 4. Phtmicrgraphs f virgin dark-brwn hair submitted t dryer heat (final magnificatin 480x, bright-field and transmitted light). A,B: Fibers after 50 min f heat expsure. C,D: Fibers after 110 min f heat expsure. Nte that the medulla disappears prgressively with dryer heat expsure.

446 JOURNAL OF COSMETIC SCIENCE became darker after ht-plate treatments. This als means that the main structural changes happen in the cuticle cells, as they are respnsible fr hair lightness prperties (9). When silicne fluid is applied t the tresses, quite smaller clr changes are bserved in DE*, suggesting that it preserves clr prperties (see Table II). Hwever, a cmparisn between the clr reference values (ftntes in Tables I and II) disclses an initial darkening effect f the hair tresses treated with silicne fluid, in much the same amunt as the ht-plate effect. Silicne fluids are well knwn as cnditining agents that smth and increase hair glss (11,12). This is expected t cause clr changes, but darkening is expected nly if the fluid penetrates the fibers (9). EFFECT OF DRYER HEAT These experiments were dne t bserve if the heat f a cmmn hairdryer had any effect n hair clr prperties. Table III shws n significant values fr the ttal clr difference parameter, DE*, meaning that hair expsure t dryer heat causes n appreciable changes in hair clr, althugh cmparisn amng hair sample phtmicrgraphs befre and after expsure t the dryer shws marked differences in the hair medulla. Figure 3 cnsists f phtmicrgraphs (Figures 3A, 3B) and a transmissin electrn micrgraph (Figure 3C) f cntrl hair samples with the hair medulla in detail. The medulla is bserved in 50% (rt regin) and 75% (tip regin) f the cntrl fibers, despite being fragmented and with variable thickness. A crss-sectin verview f a medulla is seen in Figure 3C, revealing its nn-fibrillar cnstitutin in ppsitin t the macrfibrills f the crtical cells. Figure 4 shws phtmicrgraphs f the rt regins f dryer-expsed, hair samples. In Figures 4A and 4B, hair fibers were expsed t dryer heat fr 50 min, and in Figures 4C and 4D fr 110 min. We fund the medulla nly in 33% and 20% f the hair fibers, respectively. Surprisingly, this pints t a labile behavir f medullar cells when expsed t the intermediate heating f 60øC prduced by the hairdryer. It seems that the disappearance f the medulla has n effect n the clr prperties in this hair sample. CONCLUSIONS We have shwn that tne-up dyeing frmulatins are effective in clr prductin and in dye persistence in the hair, as are the lng-lasting nes. On the ther hand, the clr change btained after hair dyeing depends n the previus cleansing treatment f the hair. Hair clr prperties are significantly affected by hair histry and by expsure t high temperature (curling irns). Clr changes are due mainly t a lightness increase, DL*, and can be masked using silicne fluids in hair pretreatment. On the ther hand, clr parameters are insensitive t sme drastic structural changes, such as lss f an bservable medulla, prmted by expsure t rdinary hairdryer temperatures. ACKNOWLEDGMENTS The authrs thank FAPESP (grant 01/14161-9 and grant 02/12541-1), CAPES, and CNPq fr financial supprt.

EFFECT OF DYEING AND HEAT ON HAIR COLOR 447 REFERENCES (1) N. A. Barnict and M. S.C. Birbeck, "The Electrn Micrscpy f Human Melancytes and Melanin Granules," in Bilgy f Hair Grwth (Academic Press, New Yrk, 1958), pp. 239-252. (2) R. McMullen and J. Jachwicz, Thermal degradatin f hair. I. Effect f curling irns,j. Csmet. Sd., 49, 223-244 (1998). (3) R. Alexander-Kartz and R. G. Barrerra, Surface crrelatin effects n glss, J. P/ym. Sci. B: P/ym. Phys., 36, 1321-1334 (1998). (4) ASTM D 523-85 t, Standard test methd fr specular glss, pp. 52-56 (1985). (5) G. Sauremann, U. Hppe, R. Lunderstadt, and B. Schubert, Measurement f the surface prfile f human hair by surface prfilmetry, J. Csmet. ScJ., 39, 27-42 (1988). (6) I.J. Kaplin, A. Schwan, and H. Zahn, Effects f csmetic treatment n the ultrastructure f hair, Csmet. Tilerr., 97, 22-27 (1982). (7) J. Guld and R.L. Sneath, Electrn micrscpy image analysis: Quantificatin f ultrastructural changes in hair fiber crs sectins as a result f csmetic treatment, J. Sc Csmet. Chem., 26, 53-59 (1985). (8) C. R. Rbbins, Chemical and Physical Behavir f Human Hair, (Springer-Verlag, New Yrk, 1994). (9) C. Scanavez, M. Silveira, and I. Jekes, Human hair: Clr changes caused by daily care damages n ultra-structure, Cl. Surf B: Biinterfaces, 28, 39-52 (2003). (10) C. Scanavez, M. Zega, A. Barbsa, and I. Jekes, Measurement f hair luster by diffuse reflectance spectrphtmetry, J. Csmet. Sci., 51, 289-302 (2000). (11) M. A. Berthianne, J. H. Merrifield, and D. A. Ricci, Effects f silicn pretreatment n xidative hair damage,j. Sc. Csmet. Chem., 46, 231-245 (1995). (12) K. Yakagi, Silicnes as cnditining agents in shamps,j. Sc Csmet. Chem., 43, 275-284 (1992).