Surface reactions of adhesives on dentin G. Eliades G. Palaghias G. Vougiouklakis Research Center for Biomaterials 15 Smirnis Str. 165 62 Terpsithea Greece Received November 3, 1989 Accepted April 11, 1990 Dent Mater 6:208-216, July, 1990

Abstract-The purpose of this study was to characterize changes in the surface chemistry of dentin following various adhesive treatments. The coronal parts of sound freshly extracted third molars were cross-sectioned over the pulp chambers, each producing a pair of dentin samples which were polished to 600 grit and cleaned with 3% H202. The.first sample of each pair was used as a control, while the second one was subjected to one of the following adhesive treatments: (a) Gluma Cleanser, (b) Tenure Conditioner, (c) Scotchprep, (d) Gluma Cleansed Gluma Primer, (el Tenure Conditioned Tenure Solution A&B, or (f) Scotchprep/ Scotchbond 2 Adhesive. The treated samples paired with their respective controls were studied by small-area ESCA spectroscopy. Three areas of 1.0 mm in diameter randomly chosen on each sample were analyzed by survey and Cls, 01s, Nls high-resolution spectra. The samples from groups d, e, and f were additionally subjected to argon-ion-depth profiling of the uppermost 2-nm layer at 0.5-nm intervals. According to the results, treatment modes a, b, and c caused the reduction of carbonates and increased the -NH/NH~ ratio. Treatments a and c increased the alcohol groups, while treatments b and c increased the carbonyl and ether groups. All these changes were in comparison to the reference dentin specimens. Dentin treatment with d, e, and f induced a complex in depth distribution of the C, N, 0 binding states. The energy shifts detected do not indicate primary bonding of the tested adhesives to the dental substrate.

urrent research on dental adhesives is focused on the surface alterations of dental tissues subjected to adhesive treatments. Surface characterization is of critical importance for the study of surface interactions occurring on dental tissues, because it provides unique information for the elemental concentration, molecular environment, and oxidation level of the substrate. Thus, a better understanding of the bonding mechanisms involved can be obtained. While considerable work has been conducted on bond strength, microleakage, and fracture topography of dentin adhesives, there are limited data available on chemical modificaUons induced on the substrate. Some alterations in the elemental concentration of dentin treated with dental adhesives have been reported recently (Williams and Williams, 1988; Thompson et al., 1989). However, these findings were not associated with specific surface functional compounds. The purpose of this study was to characterize changes on the surface chemical structure of dentin following various adhesive treatments.

C

MATERIALS AND METHODS The coronal p a r t of each sound, freshly extracted third molar was cross-sectioned over the pulp chambers, providing a pair of samples which were polished to 600 grit by use of wet silicon carbide papers and were cleaned with 3% hydrogen peroxide. The first sample of each pair was used as a control, while the second was subjected to one of the following treatments: (a) Scotchprep, (b) Gluma Cleanser, (c) Tenure Conditioning Solution, (d) Scotchprep + Scotchbond 2 Adhesive, (e) Gluma Cleanser + Gluma Primer, or (f) Tenure Conditioning Solution + Tenure Solution A and B. The batch numbers of the tested adhesives are shown in Table 1. Each treatment was performed according to the

208 ELIADES et al./ESCA STUDY OF ADHESIVES ON DENTIN

manufacturer's instructions at a constant temperature of 37_+ I°C. Triple-distilled water was used as an intermediate rinsing solution wherever it was suggested. The treated samples paired with their respective controls were examined by electron spectroscopy for chemical analysis (ESCA) on a Perkin-Elmer 5400 ESCA system (Perkin-Elmer, Physical Electronics Division, E d e n P r a i r i e , MN) equipped with a variable-aperture lens capable of analyzing small areas. Three areas of 1.0 mm in diameter were randomly chosen and were analyzed by survey and high-resolution spectra employing a Mg K(~ anode at 250W and 45 ° take-off angle. Assuming 2.5 nm as a mean free path for Cls photo-electrons excited by Mg Ks x-rays (Clark and Thomas, 1977), the mean sampling depth relative to Cls for each acquisition was calculated as 1.6 nm. From the survey spectra, the qualitative and quantitative elemental composition and state of the uppermost layer were computed. The binding states of each element were obtained by acquisition of high-resolution spectra and subsequent curve-fitting. The assignment of each curve-fit to a specific binding energy was characteristic of the compounds formed on the dentin surface. For treatments d and e, which resulted in a very thin film, depth profiling was performed after argon-ion-sputtering at 2 KV and subsequent analysis. Argon ions rastered over the I-ram-diameter area at a milling rate of 3 nm/min were used so that we could investigate the uppermost 2-nm layer at 0.5-rim intervals. At each depth, survey and high-resolution spectra were also acquired under the same conditions. For treatment f, which provides a thick adhesive layer, a different procedure was used. Data acquisition on depth profiling and survey spectra was continued until the interface was identified. Profiling was repeated again at another position until the

distance of i nm from the previously established resin-dentin interface was attained. Profiling was then suspended, and survey and high-resolution spectra were acquired. The same procedure was performed at 0.5-nm intervals extending to 1 nm beyond the interface. The plotted spectra were chargeco~Tected with the aliphatic Cls peak at 284.60 eV used as internal standard (Ratner, 1983). Following cm~vefitting of each peak, the resulting binding energies were assigned to the corresponding compounds based on catalogued data (Wagner et al., 1979). The statistical analysis of the results was performed by ANOVA and the Newman-Keuis multiple comparison test at a 0.05 significance level.

TABLE 1 THE TESTEDDENTALADHESIVES Adhesive SCOTCHBOND2 Scotchprep Scotchbond 2 Lightcure Adhesive GLUMA SYSTEM Gluma Cleanser Gluma Primer TENURE SYSTEM Conditioner Solution A Solution B FILE: _25 Sanple RI SCPJ_EFACT~, OFFSE1=11.808,

I

10

Manufacturer

7E7P 7E3P

3M Dental Products Co. St. Paul, MN, USA

4108S 4104S

Bayer Dental, Leverkusen, FRG

461002 462002

Den-Mat Corporation Santa Maria, CA, USA

8/24/88 ANGLE:45 de9 ~Q 1II4E=8.25 ein

ESCA~

I

I

I

I

0.693 k c/s ~

I

I

I

I

EIf_RGY=89.450 eV 149 250 N

I

I

I'

I

I

I

I

I

I

I

I

I -~

9 8

RESULTS

Two representative ESCA survey spectra of reference dentin and dentin t r e a t e d with Scotchprep are shown in Fig. 1. Scotchprep reduced Ca2p and P2p peaks, while it increased Nls, Zn2p, and Si2p peaks. The 01s/Cls ratio did not change. In Fig. 2 an example of high-resolution spectra is given. On dentin treated with Tenure, the Nls highresolution curve-fitted spectrum revealed a shift in the binding energies and an additional nitrogen phase compared with the phases appearing on reference dentin. The relative atomic concentration ratios of C/O, C/N, Ca/C, and Ca/P after dentin treatment with cleansinn and conditioning solutions are presented in Table 2. All the treatmerits reduced C/N, Ca/C, and Ca/P ratios compared with reference dentin. An increased C/O ratio was noted for Gluma Cleanser and Tenure Conditioner. The relative functional group concentrations and the ESCA binding energies corresponding to high-resolution analysis of cleansed and conditioned dentin surfaces are shown in Table 3. The statistically significant differences obtained from the multiple comparison test are presented in Table 4. Cls high-resolution analysis revealed that Scotchprep increased the surface alcohol groups (C-OH) over carbonyl (C = O) and aliphatic ( - CH) groups, compared with reference dentin.

Batch No.

7 G

o

5

4

-

N

3 2

1 0 1100.0

I

I

b

990.0

I

I

I

I

770.0

880.8

I

I

I

I

t

I

f~oO.O 550.0 4~.0 BINDING ENEMY, eV

f

I

3~.0

I

2~.0

110.0

O.O

ESCfi SURVEY 8/21/88 tiNGLE:45 deg ACQ1IME:18.34 ~in FILE: El _I Sac,pie S12 SCALEFEI~, OFFSET= 9.153, 0.500 k c/s PtiSSENERGY=89.450 eV M9 250 g

1{}

8

~

I

I

I

I

I

I

,

I

I

]

I

I

I

P

I

~

I

I

I

\ ~c o . ~

7

c.

~

~,J 4

~

t 1100.0

990.0

880.0

770.0

GGO.O 550.0 440.0 BINOIN6 E~RGY. eV

330.0

220.0

110.0

O.O

Fig. 1. ESCA survey spectra of reference dentin (a, lop) and dentin treated with Scotchprep (b, bottom).

Gluma Cleanser d e m o n s t r a t e d a substantial increase in acid (CO0) and carbonyl (C=O) functional groups,

while Tenure Conditioner exhibited increased concentration of alcohol and carbonyl groups.

Dental Mate~'ials/July 1990

209

(SOl ~ FIT 8/24/88 NIGLE:45 dq IEQ 13)[:13.37 nin FILE: Curve_Fit SampleRI $CnLEFRCIOR. OFFS~: 0.138, O.OOOkc/s PRSSE]i]~'Y=-35.758eV 1(9 250N I

I

I

I

I

I

I

9 8 7

I

'

5 ,

A

Y,

3 2

1 e

~"~-

411.5

.

'*"'"

.

,

408.5

410.0

.

,

485.5 404.0 402.5 I l I ~ I ~ EIQGY, eV

407.0

3~.5

401.0

3~.5

3~.e

ESCIICIII~F]I ~ IVlGIJE=45dq lIC01]NE=7.02nln Cur~t_Fit $ a ; l e 12. 2 m 1,Ate'~l SClll.(FllCln= 0.271lkc/s, QFFSL~= O.UOkc/s FnssE]Ig31G~-36.750eV Jig 300N

m

I

I

I

I

I

I

I

I

I

I

I

-I

I

"

'

;

;

;

'

$ 8 7

6

I!, 3

1

e

~z'k.

45.e

4o4.0

~

403.0

,m.e

401.0

400.e

3ss.e

sse.e

~7.e

~.e

1~4)ING ENERGY, Fig. 2. Nls high-resolution ESCA spectra of reference dentin (a, top) and dentin treated with Tenure Conditioner/Solution A&B at 2.0 nm nominal depth (b, bottom).

3m.e

TABLE 2

THE ATOMIC CONCENTRATIONRATIOS OF DENTIN SURFACETREATEDWITH CLEANSERSAND CONDITIONERS, UPPERMOSTLAYER (DEPTH, 0.0 nm) Adhesive Treatment Scotchprep

C/O (_+0:~)"

Ratio (Mean _+ 1 SD) C/N Ca/C 5.51 . 0.07 1 (___1.12)

(-0.01) /

0.06 / 2.64 5.72 (_0.01)| [(±0.65) (_+1.92) 0.02 [ Tenure Conditioner L 2.49 5.25 (_+0.01) (_+0.99) (_+2.12) 0.56 Reference Dentin 1.68 11.71 (+_0.01) (_+0.32 (___2.21) Bars indicate values of no statistical si nificant difference at p

20 10

0.0

i

i

i

0.5

1.0

1.5

2.0

DEPTH PROFILING ( nm )

SCOTCHPREP / SC 2 RESIN Cls BINDING STATES 8O DENTIN

70 Z

_o I---

6o

¢v-

~

50

O -CH -CH2-OH

Z

o

4O

o o

30

Z

•

[ ] -COO-CH2

MJ p~..I

,-

-COO M+

20 10

I

0

-1.0

-0.5

0.0

0.5

1.0

DEPTH PROFILING ( nm )

Fig. 4. Relative concentrations of Cls binding states on dentin as a function of depth profiling. (a, top) Gluma Cleanser/Gluma Primer. (b, middle) Tenure Conditioner/Tenure Solution A&B. (c, bottom) Scotchprep/ Scotchbond 2 adhesive.

Dental Materials/J~dy 1990 2~3

GLUMA CLEANSER / PRIMER Ols BINDING STATES 80 =~

_o

70

~

60

~ Z

50

oZ o

40

o

30

~-

20

"

10

(l~ -CO3 Si02 BB

C-O-C

[]

-NH2 HSO3.

mm

m

0.0

0.5

1.0

1.5

2.0

DEPTH PROFILING ( nm )

TENURE CONDITIONER / SOLUTION A & B Ols BINDING STATES

80

70

so

~1~ -03

~,.=,

so

=o

4o

m" MMA-COO-

o°

30

r-] -NH2 HSO3

SiO2

2O ,o 0 ~

0.0

0.5

V

V

V

1.0

1.5

2.0

DEPTH PROFILING ( nm ) SCOTCHPREP / SC 2 RESIN Ols BINDING STATES _~ z

ADHESIVE

I J

DENTIN

100

2

•~

80

A

mCO3

m o oz

60

~ •

Si02 MMA-COO-

,.

40

[]

C-O-C

~wl

20

Z

O

3

-1.0

-0.5

0.0

0.5

1.0

DEPTH PROFILING ( nm ) Fig. 5. Relative concentrations of 01s binding states on dentin as a function of depth profiling. (a, top) Gluma Cleanser/Glum Primer. (b, middle) Tenure ConditionerRenure Solution A&B. (c, bottom) Scolchprep/ Scotchbond 2 adhesive.

214

E L I A D E S et al./ESCA S T U D Y OF A D H E S I V E S ON D E N T I N

crease in carbonates. EDTA treatment removes the dentin smear layer (Asmussen and Munksgaard, 1985). A possible reaction mechanism of Gluma Cleanser may involve chelation with dentin calcium, dissolution of the soluble NCPs, reduction of the electrostatic forces among the smear particles, and finally removal of the smear layer by water-rinsing. Probing the nitrogen molecular environment of dentin after EDTA treatment revealed that nitrogen is assigned to NH3OH- groups and some amino acid zwitterions. The dipolar form of amino and carboxyl terminal ends in zwitterions may be explained by the surface effect of the neutral aqueous EDTA solution. However, the greater concentration of the hydrated -NH2 ÷ groups indicates that amino side-chains are the active nitrogen groups of dentin after EDTA treatment. The possibility that the detected nitrogen signals originated from EDTA residues is limited, since the nitrogen state in EDTA does not correspond to the recorded binding energies. Tenure Conditioner increased the alcohol, carbonyl, and ether groups of dentin, while it reduced carbonates due to the low pH of the aluminum oxalate solution. There was no substantial difference in the NH/ -NH2 binding states compared with the reference. However, a well-defined energy region of NH4N03 was revealed which can be explained by the assumption that nitrate radicals are incorporated in the oxalate solution, as has been proposed by Bowen (1986). Nitrates may selectively attack the e-amino groups of lysine and hydroxylysine, which are strong bases forming ammonium compounds. Surprisingly, no aluminum was d e t e c t e d in the s u r v e y spectra. This finding suggests that either aluminum diffused deeply in the substrate projecting the oxalate ends, which may react with active dentin sites, or it was washed away during rinsing. No scission of oxalares was detected in this stage of the treatment. The three adhesive systems included in this study require the subsequent application of an adhesive compound over the pretreated dentin. According to the manufacturers,

Scotchbond 2 light-cure adhesive is composed of a BisGMA/2-HEMA m o n o m e r , Gluma P r i m e r is an aqueous solution of 2-HEMA/glutaraldehyde, and Tenure Solutions A&B contain NTG-GMA and PMDM in acetone. Both Gluma and Tenure provide a very thin film in contrast to Scotchbond 2 adhesive, which is applied in a 100-txm layer. The only significant difference recorded in the atomic c o n c e n t r a t i o n r a t i os as a function of depth for Gluma and Tenure was the lower C/N ratio for the former. Both these treatments remove the smear layer. However, aluminum oxalate modifies the dentin surface. On such a surface, it has been p r o p o s e d t h a t NTG- forms charged complexes which initiate PMDM polymel%ation (Bowen et al., 1982). This treatment reduces the total amount of amino sites for bonding, compared with EDTA. The rapidly decreasing ratio of Ca/ P found at the resin side of Scotchbond 2 Adhesive compared with the same ratio at the dentin side suggests that increased calcium concent r a t i o n was p r e s e n t . A possible explanation might be that Scotchprep causes calcium to be released from the smear layer. Thereafter, Scotchprep copolymerizes with the adhesive layer. The lower C/O ratio at the adhesive side also implies that the increased concentration of oxygen phases d etected near dentin probably arises from the orientation of hydrophylic moieties toward dentin. A complex depth profile of Cls and Ols binding states was detected in Scotchbond 2 Adhesive treatment. The reduction of the aliphatic carbon may apparently be attributed to the decrease of hydrophobic groups at the interface. Alcohol and carboxyl compounds were modified in such a way that the relative concentrations of these groups were inversely related. A reasonable explanation might be that some maleic groups were esterified to -OH groups of organic molecules, while others reacted with -Ca. The higher concentration of the organic esters at the adhesive side supports the hypothesis that a reaction between maleic acid and 2HEMA proceeds. Furthermore, an increased concentration of alcohol

GLUMA CLEANSER / PRIMER Nls BINDING STATES

~ z _o < ~z ,,, o z oo "' > P--J IJJ "

100 90 80 70 60

-NH/-NH2 1:2

50

0

-N