Nuclear Instruments and Methods in Physics Research B 316 (2013) 192–197

Contents lists available at ScienceDirect

Nuclear Instruments and Methods in Physics Research B

journal homepage: www.elsevier .com/locate /n imb

Strain buildup in GaAs due to 100 MeV Ag ion irradiation

0168-583X/$ - see front matter � 2013 Elsevier B.V. All rights reserved.http://dx.doi.org/10.1016/j.nimb.2013.09.010

⇑ Corresponding author. Tel.: +91 3222283856.E-mail address: anushree@phy.iitkgp.ernet.in (A. Roy).

Shramana Mishra a, Sudipta Bhaumik a, Jaya Kumar Panda a, Sunil Ojha b, Achintya Dhar a, D. Kabiraj b,Anushree Roy a,⇑a Department of Physics and Meteorology, Indian Institute of Technology, Kharagpur 721 302, Indiab Inter-University Accelerator Center, Aruna Asaf Ali Marg, New Delhi 110 067, India

a r t i c l e i n f o a b s t r a c t

Article history:Received 28 June 2013Received in revised form 26 August 2013Available online 5 October 2013

Keywords:Ion irradiationGaAsRamanHRXRD

The formation of strained layers and a non-monotonic evolution of strain in high energy (100 MeV) silverion (Ag7+) irradiated undoped semi-insulating GaAs are observed and analyzed using Raman scatteringand high resolution X-ray diffraction (HRXRD) measurements. At low fluence, compressively strained lay-ers are formed, whereas, with increase in fluence both compressive and tensile strains appear as observedfrom HRXRD measurements. Further, at low fluence, the change in compressive strain with increase influence is found to be sharper than what is observed at higher fluence, thereby suggesting a critical flu-ence value, beyond which there is a simultaneous generation and annihilation of vacancy type defects.The initial blue shift and subsequent relative red shift beyond above critical fluence in the Raman peakalso qualitatively reveal non-monotonic evolution of strain in this case. Finally, we demonstrate the sen-sitivity of Raman spectroscopy in detecting the decrease in lattice ordering in the crystal in the low flu-ence regime, below the detection limit of Rutherford back-scattering channeling (c-RBS) measurements.

� 2013 Elsevier B.V. All rights reserved.

1. Introduction

Group III–V semiconductors find wide scale applications inmodern technology. The optical and the structural properties ofthe semiconductors play a crucial role in device engineering. Fortuning the characteristics of an as-grown semiconductor, a varietyof techniques are used. High energy ion irradiation is one such tool[1–4]. When penetrating a solid, swift heavy ions (SHI) lose theirenergy predominantly through inelastic interactions with thetarget electrons. The process is fast (the time-scale is of the orderof few pico-seconds) and is confined in a space of a narrow cylin-drical region (of diameter few nanometers and of length severalmicro-meters). SHI modify materials structurally, chemically orelectrically along their path of motion [1–8]. This promptedresearchers to tailor material modifications by SHI irradiation inorder to achieve novel functionalities.

There are quite a few reports in the literature, which character-ize the structural evolution in SHI irradiated semiconductors. It isobserved that in SHI irradiated III–V binary semiconductors; crys-talline to amorphous phase transformation occurs [4,7,8]. Smallangle X-ray scattering (SAXS) [5] and infra-red absorption spec-troscopy [6] reveal that SHI irradiation results in the breaking of

atomic bonds. It has been shown that a nanometer sized damagedzones are formed by single ion impact during SHI irradiation [7,8].Various theoretical models are available in the literature whichexplains the synergy of electronic and nuclear energy loss (ee anden, respectively) during the process of irradiation in a semiconduc-tor [6,9]. The newer and finer experiments on SHI irradiatedsemiconductors would enable the researcher to refine the existingmodels for better understanding of the microscopic nature of theprocess.

In this study, we have used room temperature Raman scatter-ing, high resolution X-ray diffraction (HRXRD) and Rutherford backscattering channeling (c-RBS) measurements to understand themicroscopic nature of the defects, which are formed in GaAs dueto 100 MeV silver ion (Ag7+) irradiation. We exploit the sensitivityof Raman spectroscopy to study the structural evolution at lowerfluences (below the detection limit of c-RBS) of irradiation. At high-er fluence range, the irradiation induced defect-disorder in thematerial, as obtained from Raman measurement, could be corre-lated with the same as revealed from c-RBS measurements. Wehave observed an evolution of non-monotonic compressive strainin the irradiated sample by HRXRD measurements. The change inthe slope of the plot of compressive strain in SHI irradiated GaAsvs. ion fluence indicates that beyond a critical fluence of irradia-tion, there is a simultaneous generation and annihilation of defectstates. Furthermore, we show that along with the compressivestrain, the tensile strain develops beyond this critical fluence ofirradiation.

S. Mishra et al. / Nuclear Instruments and Methods in Physics Research B 316 (2013) 192–197 193

2. Experimental details

High resistive (�106 X cm) undoped semi-insulating GaAs(SI-GaAs) single crystal wafers of orientation (001) and thickness�150 lm, grown by liquid encapsulated Czochralski (LEC) tech-nique, were subjected to a 100 MeV Ag7+ irradiation in the fluencerange of 6 � 1011– 6 � 1013 ions/cm2 using 15UD Pelletron at IUAC,New Delhi. In case of 100 MeV Ag ions irradiation in GaAs, en is�0.11 keV/nm, whereas ee is �17 keV/nm, as calculated usingMonte–Carlo TRIM (Transport of Ions in Matter) simulation [10].It is to be noted that in our case the electron energy loss is less thanthe threshold energy (ee � 33 keV/nm) required for track formationin GaAs [2]. The range of the ions in the material is �10.9 lm.

HRXRD measurements were carried out using X-pert pro MRDdiffractometer with CuKa1 radiation. Diffraction patterns weremeasured at the GaAs (004) reflection, which is sensitive to thestrain in the vertical layer structure. The X-ray penetration depthin this case is larger than 5 lm [11]. The c-RBS measurements wereperformed at IUAC, New Delhi using 1.7MV Pelletron facility.2 MeV Helium (He) was used for c-RBS experiment, where back-scattered He ions were detected at an angle of 160�. A four-axisgoniometer was used to align the crystals with respect to thecollimated incoming He ions. Micro-Raman spectra were collectedin backscattering geometry using 488 nm argon ion laser as anexcitation source with a power of 10 mW. The incident beamwas focused on the sample to a �2 lm diameter spot using a50� objective lens. The spectrometer was equipped with a singlemonochromator (model TRIAX 550, make JY, France) with an edgefilter and a charge coupled device (CCD) detector.

3. Results and discussion

3.1. Strain in the lattice

Fig. 1 shows GaAs (004) HRXRD curves for undoped SI-GaAs irra-diated with high energy Ag ions at different fluences. The main peakcentered at 33.03� is from the undamaged portion of the sample. Theabsence of diffraction fringes may due to the building up ofnon-uniform strain originating from high energy heavy ion irradia-tion. We observe that for the lowest fluence of irradiation (6 � 1011

32.9 33.0 33.1 33.2 33.30.1

1

10

100

1000

1012 1013 1014

-0.2

0.0

0.2

0.4

X1000

Inte

nsity

(arb

. uni

ts)

Ω (deg)

Stra

in(%

)

Fluence(ions/cm2 )

Fig. 1. HRXRD pattern of undoped SI-GaAs wafers, irradiated with Ag+ ions fordifferent fluences (6 � 1011 ions/cm2 (black), 3 � 1012 ions/cm2 (red), 1 � 1013 ions/cm2 (green), 3 � 1013 ions/cm2 (blue) and 6 � 1013 ions/cm2 (cyan)). Inset showsthe compressive and tensile strain percentage obtained from HRXRD analysis. (Forinterpretation of the references to color in this figure legend, the reader is referredto the web version of this article.)

ions/cm2) the line-shape of the diffraction pattern is Gaussian with-out any extra peak, suggesting absence of any strain in GaAs on irra-diation at this level. With increase in fluence, the main peak isbroadened and a new peak appears at lower angle. These two newcharacteristics of diffraction patterns are signatures of the increasein disorder of the bulk material and the formation of compressivelystrained layer with a lattice constant larger than that of unirradiatedcrystal respectively. With further increase in fluence, diffraction fea-tures both at lower and higher angles appear. The diffraction peak atlower angles can be attributed to the irradiated regions, which has alarger lattice constant than the substrate. Similarly, the one at higherangle may be originating from some other regions of smaller latticeconstant than the substrate. It is to be noted that these additionaldiffraction peaks are appreciably broad, which indicates the distri-bution of strain in the material. Here we would like to mention thatas the projected range of ions and the penetration depth of X-ray inthe samples are�10.9 lm and�5 lm respectively, there is possibil-ity of HRXRD to probe the disorder due to both electronic and nucle-ar energy loss (ee and en) of the incident ions in the material.However, as ee � en (refer to Section 2), the former is expected tohave the dominant effect in the sample, which we probe by theabove measurements.

From each diffraction pattern, an approximate measure of themaximum effective perpendicular strain emax was obtained fromthe angular separation between the Bragg peak of the unstrainedand the subsidiary strained peaks from the expression [12]:

emax ¼ � cotðhBÞDh0; ð1Þ

where hB is the Bragg angle for unstrained GaAs and Dh0 is the larg-est peak separation from the unstrained peak. emax is positive ornegative depending on whether there is a net effective compressiveor tensile strain in the lattice in the perpendicular direction. Thevalues of percentage of emax as obtained from this analysis areshown in the inset of Fig. 1 (filled circles) for different fluences ofirradiation.

A large number of reports are available in literature, wherecompressive and tensile strains were observed independently onion irradiation in III–V semiconductors [11–19]. For low energy(keV) ion-implantation [11,12] or high energy light ion (HELI) irra-diation [13–17], only compressive strain was observed. The originof compressive strain in irradiated GaAs is explained in view ofgeneration of vacancy-type defects during irradiation [13–15]. Ina vacancy-type defect, a missing atom in the lattice causes inwarddisplacements of the neighboring atoms. The magnitude of the dis-placement decreases as the neighboring atoms are away from thecenter of the vacancy. Thus, unequal-distance displacements ofthe neighboring shell atoms result in an expansion of the latticespacing [13–15]. Considering diffusion of defects during irradia-tion, it has been suggested that the predominant defects in thenear-surface disorder region in MeV-ion-irradiated GaAs arevacancies and their complexes [15].

The tensile strain had been reported in InP on SHI irradiation[18,19]. Interstitials (i.e., displaced atoms) in a damaged crystalproduce an opposite displacement direction of the surroundingatoms compared to what is observed for a vacancy type defectsin the system. However, for this particular case compressive strainwas not observed. The appearance of both compressive and tensilestain in irradiated samples, as observed by us, is most likely due tothe simultaneous formation of vacancy and interstitial defects inAg irradiated GaAs beyond a certain fluence of irradiation. Theappearance of both types of defects in Ag ion irradiated GaAsmay be due to the fact that the mass of Ag ions and energy of irra-diation, used by us, are much higher than those ions and energyreported in the above references [13–16,18,19]. Here, the tensilestrain appears at the fluence of 1 � 1013 ions/cm2. At this fluence

194 S. Mishra et al. / Nuclear Instruments and Methods in Physics Research B 316 (2013) 192–197

the compressive strain is observed to be 10 times higher than thetensile strain.

3.2. Non-monotonic evolution of the compressive strain from HRXRDmeasurements

Fig. 2 shows that the compressive strain, as obtained fromHRXRD measurements, increases non-monotonically with increasein fluence. The rate of evolution of the strain with increase in ionfluence reduces by an order of magnitude beyond a certain fluenceof irradiation (1 � 1013 ions/cm2). We define this as the criticalfluence. As mentioned earlier, due to the increase in defectconcentration, ion irradiation induces a positive effective perpen-dicular lattice strain with an increase in ion fluence [12]. However,beyond the critical fluence of irradiation the rate of increase of thecompressive strain with ion fluence decreases followed by an onsetof simultaneous generation and annihilation of defect states in thesystem. At this point, some of the new vacancy type defects, gen-erated by irradiated ions, are annihilated by the process of recom-bination. There are few other reports in the literature, wheresimilar behavior of strain with fluence had been observed for dif-ferent ions [13–16]. In the case of MeV ion irradiated GaAs, fromthe model of single ion-lattice collisions [13], it has been concludedthat a combination of collision damage and electronic ionizationcan set in an equilibrium of defect population in the system andhence, a saturation of the surface lattice strain at a certain fluenceof irradiation. Next we try to understand the role of defects pro-duced by irradiation in determining the critical fluence i.e, thepoint of kink in the strain vs. fluence plot in Fig. 2 mentionedabove. We have estimated the displacement per atom (DPA) [4],at the fluence where the slope of the strain vs. fluence plot changes,in references [13–16] for GaAs irradiated with ions of various massand energy and also in our case (Table 1). We find that in our casethe DPA at the point of kink (critical fluence) in the plot is almostten times less than the values obtained for other ions (in abovereferences). Also the strain at this point in 100 MeV Ag ion irradi-ated GaAs is 0.23%, which is much less than the value of the sameobserved for other irradiating ions in these references. Table 1 alsoincludes the value of ee for all other ions in GaAs (as estimatedusing TRIM simulation). It is to be noted that the value of ee forour case is an order of magnitude higher than the same for otherenergetic ions. Thus, it is not very surprising that in 100 MeV Agion irradiated GaAs, the critical fluence for observed modulationof strain is lower than what is observed for other ions in theliterature [13–15].

0 2 4 6-0.1

0.0

0.1

0.2

0.3

0.4

x1013

Fluence (ions/cm2)

stra

in (%

)

Fig. 2. The variation of the compressive lattice strain measured by HRXRD (d) withirradiation fluence. The dashed lines are the linear fit at different fluence regime.

3.3. Lowering of lattice ordering

In addition to the strained layers, high energy ion bombardmentalso results in damage/disorder in the irradiated material. As men-tioned earlier the broadening of the main diffraction peak in Fig. 1is a signature of disorder on ion irradiation. Raman scattering is ahighly sensitive technique, which can detect the breakdown of lat-tice ordering even to a small length-scale (order of nanometers).Fig. 3 shows the room temperature Raman spectra (over the range200–350 cm�1) of undoped SI-GaAs irradiated with 6 � 1011–6 � 1013 ions/cm2. For comparison, we have included Raman spec-trum of the unirradiated GaAs at the top of the same figure. Thesharp features at 265 cm�1 and 292 cm�1 are the characteristictransverse optical [TO(C)] and longitudinal optical [LO(C)] modesof crystalline GaAs. The relatively broad structure in the irradiatedsamples centered around�250 cm�1 corresponds to the disorder in-duced/activated (D-A) mode of GaAs [20,21]. Inset of the Fig. 3 showsthe room temperature Raman spectra (over the range 275–300 cm�1) of all samples. Comparing the Raman line profile withunirradiated GaAs, we find that the irradiation results in asymmetryin the Raman line shape. Disorder of any kind in the crystal structuremanifests itself in two ways. One gives rise to the asymmetry in theRaman line shape, described by the spatial correlation model (SCM),mentioned below and the other by the appearance of disorder acti-vated broad modes in GaAs as mentioned above.

In an ideal crystal phonon is a plane wave. Hence, the wave-vec-tor (q) selection rule requires q � 0 for first order Raman scattering.However, for a disordered sample, as there is a breakdown of thelong range ordering in the crystal, the spatial correlation functionof phonon becomes finite [21] and the q � 0 selection rule isrelaxed [20–22]. Associated with this there is a finite correlationlength L for phonons. The allowed range of q in the phonondispersion curve is then proportional to 1/L. This model had beensuccessfully used to account for q-vector relaxation related tostructural disorder in alloyed semiconductors [23] and ion-damaged materials [20,21]. Assuming that the finite size of thecorrelation regions in the damaged material are spherical, theintensity of the Raman spectrum I(x) is given by [20–22],

IðxÞ ¼ I0

Z 1

0exp

�q2L2

4a2

!d3q

½x�xðqÞ�2 þ ½CB=2�2: ð2Þ

Here x(q) and CB are the phonon dispersion curve and the nat-ural line width of the corresponding bulk material respectively.The dispersion relation x(q) for GaAs is given by, x (q) =a + bcos(pq), with a = 267.8 cm�1 and b = 22.5 cm�1 and CB = 3.0 -cm�1 [20]. I0 is an arbitrary constant and ‘a’ is the lattice constant,which for GaAs crystal is �5.65 Å [20]. The integration in Eq. (2)must be performed over the entire first Brillouin zone. Sincephonon dispersion curve of GaAs is a decreasing function of thewave-vector [24], one expects a red shift and an asymmetrytowards the lower wave-number in the first order Raman line.Instead, in Fig. 3 (inset), we find blue shift in the spectra at lowerfluence of irradiation. This observation, along with the resultsobtained from HRXRD measurements, prompted us to takeinto account the additional contribution of effective strain in thespatial correlation model. Strain can lead to an additional shift inthe Raman peak position. Assuming the strain to be isotropicwithin the narrow probing depth (�400 Å [25]) of Ramanmeasurements, we have included the effect of strain in the latticein Eq. (2) as [26],

IðxÞ¼ I0�Z 1

0exp

�q2L2

4a2

!d3q

fx�½xðqÞþDxðq;LÞ�g2þ½CB=2�2: ð3Þ

Table 1Displacement per atom (DPA) for different incident ions.

Reference no. Incident ion species and energy Electronic energy loss (ee) (eV/nm) The fluence at which the rate changes (ions/cm2) DPA Strain (%)

This report 100 MeV Ag 1.6 � 104 1 � 1013 2.6 � 10�3 0.23[13,14] 15 MeV Cl 5.2 � 103 1 � 1015 1.7 � 10�1 0.4[15] 2 MeV O 1.5 � 103 2 � 1014 3.4 � 10�2 0.4[16] 1 MeV C 9.9 � 102 1 � 1014 1.4 � 10�2 0.39

200 250 300 3500

1000

2000

3000

4000

5000

(vi)

(v)(iv)

(iii)

(ii)

(i)

Inte

nsity

(arb

.uni

ts)

Raman Shift (cm-1)

275 280 285 290 295 300Raman Shift (cm-1)

Fig. 3. Room temperature Raman spectra of undoped SI-GaAs, irradiated with Ag7+

ions at different fluences ((i) un irradiated, (ii) 6 � 1011 ions/cm2, (iii) 3 � 1012 ions/cm2, (iv) 1 � 1013 ions/cm2, (v) 3 � 1013 ions/cm2 and (vi) 6 � 1013 ions/cm2). Foreach set, raw data points are shown by (+) and the solid line is the net fittedspectrum to the data points using Lorentizian function for all Raman modes andspatial correlation model (including strain) for the LO phonon. The de-convolutedcomponents for the sample irradiated at the fluence 6 � 1013 ions/cm2 are shownby dashed lines in the bottom of the figure. Inset shows the Raman spectra (over therange 275–300 cm�1) of all samples.

1011

1012

1013

1014

60

70

80

90

100

110

120

L(A0 )

Fluence (ions/cm2)

Fig. 4. The variation of the correlation length L as obtained from Raman scattering(j) with fluence of irradiation. The value of L for the unirradiated sample isindicated by a filled circle on the y-axis. The dashed lines are guide to eye.

10 11 10 12 10 13 10 14-0.1

0.0

0.1

0.2

0.3

0.4

0.5

Fluence (ions/cm2)

Dam

age

Frac

tion

(fRS)

Fig. 5. The evolution of Raman damage fraction (fRS) as a function of fluence. Thevalue of fRS for the unirradiated sample is indicated by a filled circle on the y-axis.The dashed line is a guide to the eye.

S. Mishra et al. / Nuclear Instruments and Methods in Physics Research B 316 (2013) 192–197 195

Here, Dxðq; LÞ ¼ �3cðqÞxðqÞn [26], where, c(q) is the modeGruneisen parameter and n is the measure of an effective strainin the material. The LO phonon line of all spectra in Fig. 3 are fittedwith Eq. (3) keeping L and n as free fitting parameters. For GaAscrystal c = 1.23 for zone center (q = 0) LO(C) phonon [27]. Thisvalue of c is used for all q in the first Brillouin zone. The de-convo-luted components for GaAs irradiated by Ag ions at the fluence6 � 1013 ions/cm2 are shown in the bottom of Fig. 3. Here wewould like to mention that we probe only few surface layers inRaman spectroscopy whereas X-ray penetrates deep inside thesample. Furthermore, above spectral analysis yields the net effectof both compressive and tensile strain in the system. Thus, it isnot straight forward to compare the measured strain obtained inthese two different techniques. Nonetheless, the non-monotonicchange in the strain with irradiation, as observed from HRXRDmeasurements, can be confirmed qualitatively from the initial blueand subsequent relative red shift of the LO phonon peak (dottedline in the inset of Fig. 3).

The systematic decrease in the spatial correlation length L inirradiated GaAs with increase in ion fluence, as obtained fromabove Raman analysis, is shown in Fig. 4. The decrease in L withfluence suggests a continuous decrease in lattice ordering withirradiation. We have already mentioned that there is broadeningof the main HRXRD peak on irradiation. The peak shows a

broadening from 0.005� to 0.029� over the fluence range. It sug-gests a decrease in the crystallinity of the system. Disorder in thecrystalline structure by irradiation also manifests through themode centered around �250 cm�1 of the host atoms in GaAs[20]. The systematic lowering of L is observed even in the low flu-ence regime; however, the disorder activated mode is more prom-inent at the higher fluences.

We estimate the damage fraction in the crystal by,fRS ¼ 1� ILO

ILOþITOþID�A[20], where ILO, ITO and ID�A are the integral

intensities of the LO, TO and disorder activated Raman modesrespectively. The fRS is calculated for all the samples and is plottedin Fig. 5. With the increase in ion fluence, the increase in fRS reveals

1200 1500 1800 0.00 0.01 0.02 0.03 0.04

0.0

0.1

0.2

0.3

0.4

0.5

(vi)

(v)

(iv)

(iii)

(ii)

(i)

(a) (b)

RB

S Yi

eld

(arb

. uni

ts)

Energy (keV) RBS damage fraction

RS

dam

age

frac

tion

Fig. 6. (a) RBS ion channeling spectra of undoped SI-GaAs ((i) unirradiated, (ii) 3 � 1012 ions/cm2, (iii) 6 � 1012 ions/cm2, (iv) 1 � 1013 ions/cm2, (v) 3 � 1013 ions/cm2 and (vi)6 � 1013 ions/cm2) irradiated with 100 MeV Ag7+ ions for different fluences and (b) correlation between Raman and RBS damage fraction at different ion fluences. The dashedline is a guide to the eye.

196 S. Mishra et al. / Nuclear Instruments and Methods in Physics Research B 316 (2013) 192–197

a monotonic increase in the damage in the crystal on Ag ionirradiation.

The above observations of increasing defect-disorder in GaAs byhigh energy ion irradiation have been further supported by c-RBSmeasurements. Raman scattering and c-RBS have different sensi-tivity to the types of disorder in a system. The c-RBS measurementsreflect the disordered volume due to displacement of atoms fromtheir lattice sites in a crystal [20]. On the other hand, Raman spec-troscopy can provide defect-specific information. Raman spectrameasure two aspects of the disorder, simultaneously – (i) loweringof the translational symmetry of the crystal structure and (ii) frac-tion of the disordered volume in the sample. Defects having smallvolume but different morphology (like clusters of interstitials)could influence Raman and c-RBS spectra differently. Thus, thedamage fraction in a crystal obtained by these two techniquesmay not be equal; however one can expect a correlation betweenthem. For very small concentrations of defects, they may not bedetected by c-RBS but can strongly influence Raman spectra bylowering crystal translational symmetry. In case of c-RBS, in orderto observe effective de-channeling, defects related to disorderedvolume have to be quite large, of the order of 1%. The c-RBS spectrafor unirradiated and irradiated GaAs are shown in Fig. 6(a). Thesurface peak at �1620 keV, due to scattering from Ga and Asatoms, is prominent in all spectra. Careful analysis shows damageproduced is almost negligible and far below the limit for amorph-izaion [28,29]. This supports our Raman and HRXRD measure-ments, where we do not observe large scale amorphization (it isto be noted that in both we find the signature of crystalline peakrather than fully amorphous spectrum or pattern).

RBS damage fraction (fRBS) is defined as the volume fraction ofall atoms in the ion irradiated layer, which are disorderedby irradiation. The fRBS within the irradiated layer at afixed depth for all irradiated samples is evaluated from thechanneled and random RBS spectrum using the following relation

fRBS ¼ ðY irraligned � Ypristine

aligned Þ=ðY random � Ypristinealigned Þ [7], where Ypristine

aligned and

Y irraligned are the RBS yields of the aligned spectra for unirradiated

(pristine) and irradiated samples respectively; Y random is the yieldmeasured in the random direction. We have correlated fRBS andfRS in Fig. 6(b). Earlier we have observed from Raman measure-ments that, in the low fluence regime (<6 � 1012 ions/cm2), dam-age fraction is low and increases monotonically in the highfluence regime (Fig. 5). As RBS technique is in-sensitive in lowdamage-fraction in the crystal, a good correlation between fRBS

and fRS could be obtained only for higher fluence of ion.

4. Conclusion

In this article we have studied the structural evolution due to100 MeV Ag7+ irradiation in LEC grown undoped SI-GaAs. Ramanspectral analysis reveals an increase in lattice disorder and mono-tonic increase in damage fraction with the increase in ion fluence.The damage fraction in the crystal thus obtained is correlated withthe damage fraction estimated from RBS measurements. Evolutionof compressive and tensile strain with increasing ion fluence hasbeen estimated from HRXRD measurements and further corrobo-rated by the shift in Raman LO line. The observed non-monotonicincrease in strain with increase in ion fluence has been explainedin view of defect creation and annihilation in the system duringhigh energy heavy ion irradiation.

Acknowledgement

A.R. and D.K. thank CSIR, India for financial assistance.

References

[1] A. Benyagoub, A. Audren, J. Appl. Phys. 106 (2009) 083516.[2] C.S. Schnohr, P. Kluth, R. Giulian, D.J. Llewellyn, A.P. Byrne, D.J. Cookson, M.C.

Ridgway, Phys. Rev. B 81 (2010) 075201.[3] M. Ishimaru, Y. Zhang, X. Wang, Wei-Kan Chu, W.J. Weber, J. Appl. Phys. 109

(2011) 043512.[4] W. Wesch, A. Kamarou, E. Wendler, S. Klaumunzer, Nucl. Instr. Meth. Phys. Res.

B 242 (2006) 363.[5] P. Kluth, C.S. Schnohr, O.H. Pakarinen, F. Djurabekova, D.J. Sprouster, R. Giulian,

M.C. Ridgway, A.P. Byrne, C. Trautmann, D.J. Cookson, K. Nordlund, M.Toulemonde, Phys. Rev. Lett. 101 (2008) 175503.

[6] M. Toulemonde, W.J. Weber, G. Li, V. Shutthanandan, P. Kluth, T. Yang, Y.Wang, Y. Zhang, Phys. Rev. B 83 (2011) 054106.

[7] A. Kamarou, W. Wesch, E. Wendler, A. Undisz, M. Rettenmayr, Phys. Rev. B 73(2006) 184107.

[8] F.F. Komarov, L.A. Vlasukova, V.N. Yuvchenko, T.V. Petlitzkaya, P. Zukowski,Vacuum 78 (2005) 353.

[9] H.D. Mieskes, W. Assmann, F. Gruner, H. Kucal, Z.G. Wang, M. Toulemonde,Phys. Rev. B 67 (2003) 155414.

[10] J.F. Ziegler, J.P. Biersac, U. Littmark, The Stopping and Range of Ions in Matter,Pergamon, New York, 1985.

[11] U. Zeimer, E. Nebauer, Semicond. Sci. Technol. 15 (2000) 965.[12] B.M. Paine, N.N. Hurvitz, V.S. Speriosu, J. Appl. Phys. 61 (1987) 1335.[13] C.R. Wie, T.A. Tombrello, T. Vreeland, Phys. Rev. B 33 (1986) 4083.[14] C.R. Wie, T.A. Tombrello, T. Vreeland Jr., J. Appl. Phys. 59 (1986) 3743.[15] F. Xiong, C.J. Tsai, T.A. Tombrello, T. Vreeland Jr., C.L. Schwartz, S.A. Schwartz, J.

Appl. Phys. 69 (1991) 2964.[16] G. Kuri, G. Materlik, V. Hagen, R. Wiesendanger, J. Vac. Sci. Technol. B 21 (2003)

1134.[17] D. Kabiraj, S. Ghosh, J. Appl. Phys. 109 (2011) 033701.

S. Mishra et al. / Nuclear Instruments and Methods in Physics Research B 316 (2013) 192–197 197

[18] A. Turos, J. Gaca, M. Wojcik, L. Nowicki, R. Ratajczak, R. Groetzschel, F.Eichhorn, N. Schell, Nucl. Instr. Meth. Phys. Res. B 219–220 (2004) 618.

[19] F. Eichhorn, J. Gaca, V. Heera, N. Schell, A. Turos, H. Weishart, M. Wojcik,Vacuum 78 (2005) 303.

[20] U.V. Desnica, I.D. Desnica-Frankovic, M. Ivanda, K. Furic, T.E. Haynes, Phys. Rev.B 55 (1997) 16205.

[21] K.K. Tiong, P.M. Amirtharaj, F.H. Pollak, D.E. Aspnes, Appl. Phys. Lett. 44 (1984)122.

[22] I.H. Campbell, P.M. Fauchet, Solid State Commun. 58 (1986) 739.[23] P. Parayanthal, F.H. Pollak, Phys. Rev. Lett. 52 (1984) 1822.

[24] J.S. Blakemore, J. Appl. Phys. 53 (1982) R123.[25] P.S. Pizani, C.E.M. Campos, J. Appl. Phys. 84 (1998) 6588.[26] J.E. Spanier, R.D. Robinson, F. Zhang, Siu-Wai Chan, I.P. Herman, Phys. Rev. B 64

(2001) 245407.[27] B.A. Weinstein, R. Zallen, in: M. Cardona, G. Guntherodt (Eds.), Light Scattering

in Solids, vol. IV, Springer-Verlag, Berlin, 1984, pp. 472–473.[28] W. Wesch, A. Kamarou, E. Wendler, Nucl. Instr. Meth. Phys. Res. B 225 (2004)

111.[29] R.A. Brown, J.S. Williams, J. Appl. Phys. 83 (1998) 7533.