The adsorption of oxygen on InP cleaved surfaces and its influence on schottky barrier properties

The adsorption of oxygen on InP cleaved surfaces and its influence on schottky barrier properties

Surface Science 168 (1986) 409-415 North-Holland, Amsterdam THE ADSORPTION OF O X Y G E N ON InP CLEAVED SURFACES A N D ITS INFLUENCE ON SCHOTrKY BAR...

313KB Sizes 2 Downloads 63 Views

Surface Science 168 (1986) 409-415 North-Holland, Amsterdam

THE ADSORPTION OF O X Y G E N ON InP CLEAVED SURFACES A N D ITS INFLUENCE ON SCHOTrKY BARRIER PROPERTIES A I S M A I L , A B E N B R A H I M , J M P A L A U and L L A S S A B A T I ~ R E Laboratotre d Etudea des 5ur/aces, lnterJacea et Composanta. UA 04. 0787. U S T L . Place Eugene Batatllon. 34060 Montpelher Cede~. France

Received 10 June 1985, accepted lor pubhcatlon 17 July 1985

The modlficat~on ot the electromc propemes of cleaved lnP surfaces reduced bv oxygen exposure and by Ag and AI deposits, and the effect of initial oxygenexposure on the Schottky contact propemes are studied Results show that oxygen acUon ~s h~ghenough to mask the Ag or AI submonolayer effect and to reduce the effect of the nature of the metal on the diode propemes

1. Introduction Recent results suggest that the chemistry of the interface could greatly modify and even determine the Schottky contact properties [1-7] Surface defects inducing order modification could also influence the interface formao tlon In fact, the GaAs Schottky barrier can be modulated by chemlsorptlon of gases before metalhsatlon [6] Similar results have been obtained on InP [2] Oxygen is known to modify the GaAs surface and diodes In the same way, tt produces changes m InP surface p r o p e m e s [8,9] Then, it IS mterestmg to elucidate the effect of mltml oxygen exposure on the InP Schottky contract properties. In this paper, we study, by work function and surface photovoltage topographies, the modification of the p r o p e m e s of cleaved InP surfaces reduced by oxygen exposure We show that the oxygen effect appears at low exposures (m the 1-10 L range) At 103 L, a large modification of the surface potential, espectally on p-type samples, and strong Fermi level-pinning are produced The deposition of Ag and A1 submonolayers on these surfaces does not mvolve a large modification of the pruning position Schottky diodes have been achieved by deposition of Ag and A1 dots on cleaved surfaces, before and after oxygen exposure The comparison of the diode properties shows that, in agreement with indications from surface studles, oxygen adsorption slgmficantly changes the charactenstics of the diodes

0039-6028/86/$03 50 © Elsevier Science Pubhshers B V (North-Holland Physics Publishing Division)

4 l;mail et al I Ad~orpuon o] o~ygen on hiP clemed ;urJaces

410

For Ag/InP (p-type) the barrier increases It decreases for AI/InP These results can be explained in terms of changes in the interface reactivity resulting In modifications of the surface state characteristics

2. Results Dopages of the samples are in the (2-5) x 1015 cm -~ range We put n- and p-type samples together in the same holder Cleavage and measurements have been performed under U H V conditions During oxygen exposure or measurement, the ion gauge was turned off Metal deposition was performed by an M B E - h k e technique Deposition rate was about one monolayer per minute The contact potential difference (CPD = q~sc - ~probe) and the surface photovoltage (SPV) were measured using the Kelvin method The surface photovoltage is obtained by CPD variation under strong illumination (white light, 5 x 1017 photons/cm 2 s) On certain areas the CPD and SPV topographies indicate flat-band conditions These areas, when observed by optical microscope or by M E B show no noticeable defects or imperfections We call them "'perfectly cleaved areas" as opposed to surfaces presenting noticeable surface photovoltage or

l~Pvl mV

z~qb ¢ J• 800

•O 600

~Q

~.~bp

SPV



I

2oo i



O - O - - O - O - Q--" Q - Q ~ Q .

,," A%,

-O..A

"'

]SPY

~

o cleanW

I

10*

'

10 2

oxygen

exposure

lO 3 L a n g m m r s

Fig 1 Surface photovoltage and work function variations with oxygen exposure on well-cleaved InP surfaces (Surfaces which are mirror hke are called well-cleaved surlaces ) By optical observation we do not detect defects However, a few defects can be present on these surface~

A Ismad et al / Adsorptton of oxygen on lnP cleaved surfaces

411

contact potential variations correlated with structural defects which are called badly cleaved surface Well-cleaved surfaces are surfaces presenting only a few defects and a small surface barrier On these surfaces (q~p - q~n) tS closer to the bulk Fermi-level difference (Evb. - E v b p ) and the surfaces are in flat-band or in low-depletion conditions The oxygen effect appears at low temperatures (in the 1-10 L range), a depletion layer is formed, particular on p-type samples (fig 1) When the exposure gets higher than 20 L, topographies, initially perturbed by cleavage defects, become virtually uniform Therefore q~. and ~0p are similar The surface Fermi level Evs is pruned In the 20-100 L range, C P D and SPV remain virtually constant Then, between 102 and 103 L. the work function decreases in a similar manner on n and p type while the photovoltage increases on n type, indicating contrary to the ~0. variation, an Increase of the surface barrier V~ This suggests a noticeable modification in the electromc affinity Z Table 1 gives several experimental data about q~ and ~bp variation under oxygen exposure and metal deposition These results are deduced from CPD and SPV topographies which show that the surface photovoltage on oxygenand on metal-covered areas and on areas exposed to oxygen only, are similar Ag and A1 submonolayers do not greatly modify, the photovoltage and, consequently, the initial oxygen pinning position On the other hand, ~ variations are as high as a few hundred m e V on Al-covered areas Table 2 gives the ldeahty factors and barriers of the diodes deduced from I(V) and C(V) measurements The comparison between the properties of the contacts produced on clean surfaces and after oxygen exposure shows that the barrier in A g - I n P diodes (p type) increases with the initial exposure The opposite occurs for A l - I n P M o r e o v e r the difference between q~b values obtained from I(V) and C(V) measurements IS reduced for A l - I n P diodes Note that in the latter case an important indium quantity on the surface of the dots ( = 5000 A thickness) is detected by A E S Auger spectra do not show Indium diffusion through the plots of the silvered surface

Table 1 W o r k lunctlon values (referenced to q~0. or ~0p) of n- and p-type samples after oxygen exposure and Ag or AI s u b m o n o l a y e r deposition q~0. and ~t~, are the w o r k functions of clean cleaved surfaces at flat-band c o n d m o n s , they are deduced from the variations of SPV and Aq~ resulting from the cleavage defects

¢Pn - q~.. ( m e V ) ~pp - ~p¢~,( m e V ) q~m - ~Psc ( m e V )

10~L O2

I/4 MI Ag

103 L 02 1/4MLAI + 1/4 ML Ag

10sLO~ + 1/4 ML A1

~(J - 1000 -

180 -750 -230

-~0 - 1000 ~0

-350 - 1350 -300

- 120 - 1070 -430

412

A lsrnatl et al / Advorptton o] o~.vgen on lnP cleaved sur]aces

Table 2 Schottky barrier hmghts and ldeaht) tactors for the contact~ achm~ed on clean cleaved surtaces and on oxygen-exposed cleaved surfaces Ag

10~L O, + Ag

AI

10~L O~ + AI

1 1~ 10110 1370

130 920 1155

p-type

n0deahtyfactor) ~b m ~(meV) ~b ~l~

1 10 78(I 850

1 l0 865 1000

n-type

n 0deahtv factor) q~h~ ~

1 2(I 515

1 00 44~

Ohmm at room temperature

3. Discussion T h e results p r e s e n t e d In section 2 show that surtace m o d i f i c a t i o n s are a l r e a d y n o t i c e a b l e at low e x p o s u r e s , at a tew l a n g m u i r s , the c h a n g e s in w o r k f u n c t i o n a n d surface b a r r i e r are significant and l e a d to a s t r o n g F e r m i - l e v e l p l n m n g Since q5n and SPVn r e m a i n p r a c t i c a l l y c o n s t a n t , we c o n c l u d e that the p i n n i n g occurs in the u p p e r half of the gap A t 103 L, the Evs p o s i t i o n is at ~ 0 3 e V b e l o w the c o n d u c t i o n b a n d , a p o s i t i o n v e r y close to t h a t i n d u c e d by cleavage defects This value is d e d u c e d f r o m the w o r k function e v o l u t i o n and f r o m surface p h o t o v o l t a g e analysis [9,10] In the s a m e way, e x p o s u r e reduces e l e c t r o n i c affinity m o d i f i c a t i o n s I n d e e d . a c o m p a r i s o n b e t w e e n the e v o l u t i o n of q~,l a n d ~p a n d the e v o l u t i o n d e d u c e d by t a k i n g into a c c o u n t o n l y the a b o v e E w p o s i t i o n , indicates a d e c r e a s e of • w h o s e value is ~ 0 18 e V at 10' L A c c o r d i n g l y b o t h d o n o r a n d a c c e p t o r surface states are c r e a t e d by the oxygen U s i n g a m o d e l with two discrete states, the c a l c u l a t i o n shows that the EF, p i n n i n g can be e x p l a i n e d by an a c c e p t o r level E~ at 0 9 e V a n d a d o n o r level Ed at 1 1 e V (fig 2), whmh are in a g r e e m e n t with t h o s e d e d u c e d from the unified d e f e c t m o d e l [8] A t l/) ~ L the d o n o r and a c c e p t o r state densities are similar a n d a r o u n d I0 ~ cm -2 T h e s e densities are in ag.reement with A E S i n d i c a t i o n s which show t h a t only a small o x y g e n c o n c e n t r a t i o n is p r e s e n t at the surface ( ~ 0 1 M L ) O u r m e a s u r e m e n t s d o not p e r m i t a definitive d e t e r m i n a t i o n ot the m e c h a n i s m s of the o x y g e n - s u r f a c e i n t e r a c t i o n , for e x a m p l e , the origin of the i n d u c e d states a n d X m o d i f i c a t i o n s H o w e v e r . in c o m b i n a t i o n with results from o t h e r t e c h n i q u e s [6,9,10] s o m e m o d e l s , a l t h o u g h s o m e t i m e s conflicting. can be discussed F r o m p h o t o e m i s s i o n m e a s u r e m e n t s , S p l c e r et al c o n c l u d e d [6] that the o x y g e n c h e m l s o r b s only on P a t o m s with no e v i d e n c e for b u l k oxide f o r m a tion T h e c r e a t i o n of the d o n o r and a c c e p t o r states must be e x p l a i n e d in t e r m s of d e f e c t - i n d u c e d states, i n d e p e n d e n t of the n e w surface b o n d s

A lsmall et al / Adsorption o f o ~ y g e n on l n P cleaved surfaces

413

EFs eV

,

gd

!

/ /

/,,,~Nd

1.2Na

Ea

~,-~ Nd ~ 2 Na ~

0.5

Nd

Nd - 3 Na ~ - 5 Na

Theoreticalcurves •

Experlmentalpomts

Na

0 clean

'°'° !

1011 I

10

1012

1013

I

10 2

I

10 3

1014 c m !

Langmmrs

Oxygen e x p o s u r e Fig 2 Fermi level versus oxygen exposure (expenmental points) and state densmes (theoretical curves) The ratio N j / N d decreases with oxygen exposure

(chemlsorptlon on In and/or on P sites) The exothermlc reaction between oxygen and the substrate is expected to induce defects at (or near) the interface In this case, the observed decrease of Z cannot be explained by a simple dipolar model with negative charge transfer from the semiconductor to the adsorbed oxygen a t o m s (If s o , an increase of Z would occur), even if dn adsorption mechanism, resulting in a surface dipole, would exist it must be masked by other effects, for instance relaxation of the substrate surface atoms However, the incorporation [9] of oxygen into the substrate lattice can also lead to Z modification in agreement with our observations We note that recent photoemlsslon results [10] show that spectral changes are detected at 103 L in both P 2p and In 4d core level spectra which might indicate that oxidation occurs simultaneously on the P and In sites In summary, our results show that the Schottky barrier height is substantially influenced by the initial oxygen exposure Furthermore, the pinning position induced by cleavage defects is similar to that obtained by oxygen and, moreover, the densities of the acceptor and donor states approach the same value with mcreasmg oxygen exposures Structural modifications of the surface, induced by the oxygen, may explain these observations The results of Holllnger et al [10] which indicate both In and P oxidation, support this suggestion

414

A lsmatl et al / Advorptton oJ orvgen on lnP demed ~urtaces

The Ag or A1 submonolayer depositions on a clean cleaved surface induces pinning at a position linked to the nature ot the metal Although the adsorbed oxygen concentration (<~ 0 1 ME) is less than 1/4 ML of the metal deposit, the pinning position resulting from the metal deposition and that resulting from oxygen adsorption are the same As the density of adsorbed oxygen atoms IS far lower than that of metal ones, oxygen induces many more states than the metal (if we suppose that the metal layer is uniform), the hypothesis of oxygen-induced order perturbation can explain the results In the same way, CPD and SPV results seem to agree with the hypothesis of metal cluster lormatlon, because the metal work function contribution is already effective tor 1/4 ML deposit as shown in table 1 Since the section of the Kelvin probe used is greater than the clustered area, the work function measured should be a mean value between q)m and the work function of the area not covered by the metal On the other hand, work function topographies carried out after dot deposition (5000 A thick) show that q~i,lp + o~ - q~al IS In the 0 3 eV range, while ~PI,,P + o, (])Ag lS around zero Therefore, A1 clusters must noticeably decrease the measured qhnp, while this last value should not be modified by the existence of silver clusters This conclusion is in good agreement with our experimental observations The results ot the diode studies confirm the in&cations from surface stu&es The effect of metal interaction on the pinning position is partially masked by the action of oxygen The Fermi-level of the interface approaches the pinning position induced by oxygen adsorption, similar results have been obtained on chemical InP(100) surfaces [11] On the other hand, as suggested by the reduced difference of q~B values deduced from I ( C ) and C ( V ) measurements, the interface region or the parabolic barrier deformation has been reduced, particularly with A1 However, as shown by AES analysis, indium appears in all cases on the thick A1 dots Thus, we assume that, in addition to the metal reactivity, which could influence the features of the states and thus the E~, position [12], surface order could be a p a r a m e t e r in the determination of the electronic properties of the interface -

4. Conclusion

In this paper we have given and discussed results relative to the effects of oxygen exposure and metal deposition on cleaved InP surfaces and on the Schottky barrier properties Noticeable mo&ficatlons of CPD and SPV are obtained after low oxygen exposures, especially on p-type samples A strong Fermi-level pinning is produced after a few tenths of langmuIr The deposition of an Ag or A1 submonolayer after 10 3 L oxygen exposure only slightly influences the oxygen Evs pinning Diode results confirm the surface study indications, the interface Fermi-level position approaches the pinning induced

A lsmad et al /Adsorptton oforygen on InP cleaved surfaces

415

by the initial oxygen exposure Analysis of the results leads us to explain the oxygen effect in terms of surface order modifications and to assume that the interface electromc properties are, m general, influenced by both surface reactivity and surface order modifications

References [1] L J Brlllson and C F Brucker, Appl Phys Letters 38 (1981) 784 [2] R H Wdhams, J Vacuum So Technol 18 (1981) 929 [3] V Montgomery, R H Wdhams and R R Varma, J Phys C (Sohd State Phys ) 11 (1978) 1989 [4] J Massles, F D6zaly and N T Llnh, J Vacuum So Technol 17 (1980) 1134 [5] A Ismad, J M Palau and L Lassabatere, Rev Phys Appl 19 (1984) 197 [6] W E Splcer, P W Chye, C M Garner, I Lmdau and P Planetta, Surface So 86 (1979) 763 [7] A Ismad, A Ben Brahlm, J M Palau and L Lassabat6re, Surface Scl 162 (1985) 195 [8] W E Splcer, P W Chye, P R Skeath, C Y Su and I Lmdau, J Vacuum So Technol 16 (1979) 1422 [9] W Monch, Surface So 168 (1986) 577 [10] G Holhnger, G L Hughes, F J Hlmpsel, J L Jordan, J F Morar and F Houzay, Surlace Scl 168 (1986) 617 [11] A lsmad, J M Palau and L Lassabat6re, Rev Phys Appl 19 (1984) 205 [12] A Ismad, J M Palau and L Lassabat6re, J Phys 45 (1984) 1717