Cell Processing | Passivated contacts

TOPCon technology: What exactly is

it and how mature is it in production?
Radovan Kopecek, Jan Hoß & Jan Lossen, International Solar Energy Research Center (ISC) Konstanz e.V., Germany

Abstract
Over the last few years, passivated-contact nPERT solar cells, referred gradual adaptation of existing production lines.
to as TOPCon cells, have been receiving increasing interest among solar PERC technology subsequently became, much
cell manufacturers. At the beginning, the development was mostly more quickly than anyone expected, the leading
initiated by n-type solar cell producers, but it is now also attracting solar cell technology, with the highest production
the attention of many PERC producers, who are approaching the limits capacity and the lowest cost of ownership (COO).
of their standard production lines. In this paper the situation of solar Towards the end of 2019, mono PERC production
cell production in China is summarized, and an attempt is made to capacity reached 95GWp (see Fig. 2), which
answer the question of whether passivated contacts could replace corresponds to a total annual solar cell production of
PERC technology, which will eventually reach its efficiency limit in the more than 120GW, equating to a 75% market share.
future. The most relevant passivated-contact technologies in R&D and Standard PERC cell efficiencies, however, are
production are reviewed, and the major bottlenecks impacting on a expected to reach their limits soon; scientists in the
successful industrialization are evaluated. Some parallels are outlined PV community estimate that this will happen at an
between the situation today and the evolution from homogeneous average production efficiency value of 22.5–23% [2].
Al-BSF technology to PERC in the past. Fig. 2 shows that, during the period Q2 2019 to Q3
2019, an increasing share of PERC production lines
Introduction had been upgraded to produce cells with selective
Since 2016, when LONGi began introducing their emitters, reaching a total of 75GW. This can be
low-cost Cz-Si wafers to the PV market, mc-Si interpreted as an indication that PERC producers
and homogeneous Al-BSF technologies have been are approaching the efficiency limits of this cell
rapidly losing market share, as evidenced by Fig. technology, and are squeezing out the last efficiency
1. Back then, there was stiff competition between gains from their production lines. To achieve even
passivated emitter and rear cell (pPERC) and higher efficiencies with PERC-like solar cells, new
passivated emitter, rear totally diffused (nPERT) technologies will need to be implemented.
technologies, but PERC later prevailed – mostly This raises the question of which cell concepts
because of the cheaper price of p-type wafers and will replace PERC as the leading solar cell
associated processing sequences and materials (e.g. technology of the future. Or, more specifically: how
Ag and Al pastes). An additional advantage of PERC can the voltage of low-cost industrial solar cells be
was the fact that the process sequence was closer increased towards 700mV and beyond in order to
to that for p-type standard cells, which facilitated a obtain efficiencies well above 23%? And how can

(a) (b)

Front metal grid Front ARC and passivating coating

(Al) (BSG/SiNx, SiOx/SiNx or Al2O3/SiNx)

p+ (boron) emitter

PERT/TOPCon n‐type Si
n+ (phosphorus) BSF or
oxide/n+‐doped poly‐Si

Rear metal grid (Ag Rear ARC and passivating coating (SiNx)
or Ag/Al)

Front metal grid (Ag) Front ARC and passivating coating (SiNx)

n+ (phosphorus)
emitter
PERC p‐type Si
p++ local Al BSF

Rear metal grid (Al) Rear ARC and passivating coating (Al2O3/SiNx)

Figure 1. (a) Historical data and forecast for c-Si technology market share from PV ModuleTech [1]. (b) A typical cross section of an n-type PERT solar
cell (top) and a p-type PERC solar cell (bottom).

76 www.pv-tech.org
 Passivated contacts | Cell Processing

Figure 2. LONGi’s data on Cz-Si solar cell technology market share in 2019. (PERC+SE are PERC structures that include selective emitters.)

this be achieved at acceptable costs, i.e. without the concluded that passivated contacts would need to be
addition of too many and too costly new process implemented as a next step.
steps? Fig. 3 shows that cell technologies with Solar cells with passivated contacts, in Asia often
passivated contacts (‘TOPCon’ cells) can achieve the referred to as TOPCon, a term coined by FhG ISE
efficiency goal, but at a cost that is currently not for their passivated-contact solar cell, have been
competitive with that of PERC. developed for both p-type and n-type cell concepts.
In 2019 many large PERC manufacturers, such The essential novelty with respect to conventional
as JinkoSolar and LONGi, reported at important cell technologies is that diffused or alloyed regions
Chinese PV conferences (e.g. PVSEC in Xi’an, of the cell are replaced by a stack of silicon dioxide
SNEC in Shanghai or CSPV in Shanghai) that it is and doped polysilicon (poly-Si). The replacement
becoming increasingly difficult to maintain steady of n-doped regions by oxide/poly-Si stacks with
efficiency gains with standard PERC technology. As excellent surface passivation was already achieved
contact recombination is a main limiting factor, they several years ago; for p-doped poly-Si layers, however,

WE ARE READY FOR IT!

Wet Processing
Equipment
› for M6 Wafers
› for High Efficiency Cells

700 MW
off one Wet Bench.
Now! NEW!
Carriers
by exateq

Contact us: Gerry Knoch, [email protected] exateq.de

 Cell Processing | Passivated contacts

the process is more challenging [4]. Furthermore,

it has to be taken into account that poly-Si layers
are optically highly absorptive, making full-area
application on the front side unattractive.
Most research institutes and cell manufacturers
are therefore developing n-type cell concepts with
poly-Si passivated contacts on the rear side, in
particular nPERT (Trina Solar, Jolywood, JinkoSolar,
SPIC and others) or n-interdigitated back contact
(nIBC) cells (Trina, LG electronics) structures.
Nevertheless, concepts for cell architectures based
on p-type substrates – often called polyPERC – also
exist [5]. Another alternative is p-interdigitated
back-contact (pIBC) solar cells, combining n-poly
layers with alloyed local Al contacts [6,7]; however,
such a rear-junction cell concept suffers from the
limited charge-carrier diffusion length of currently
available commercial p-type Cz-Si substrates.
At the 29th International PVSEC in 2019 in Xi’an,
an overview of the highest efficiencies obtained for Figure 3. Efficiencies, COOs and selling prices for major c-Si technologies on the PV
large solar cells in China was presented and is shown market [3].
in Fig. 4 [8]. An efficiency of 24.58% was achieved
with a TOPCon nPERT cell by Trina, and 24.03% although significant progress will still need to be
with a PERC-type cell from LONGi. While both the made in order to fulfil the remaining specifications
cells in question are assumed to use an intricately shown in Table 1. The costs for TOPCon cells must
patterned poly layer on the front side (which is not not be more than 1.2 times the costs for PERC,
suitable for cost-effective industrial production), which is not yet the case, mainly because of the
the reduced charge-carrier recombination of the conformal deposition of poly-Si in the LPCVD
TOPCon rear side, compared with a standard PERC reactors, leading to more complex processes, a lower
rear side, is clearly visible from the measured high yield and the short service life of the LPCVD quartz
Voc of the TOPCon cells. tubes (as well as because of the front and rear Ag
At the beginning of 2020, only a few companies metallization).
had started pilot production or full production Consequently, alternative inline processes – such
of TOPCon cells: examples are Trina, Jolywood, as atmospheric pressure chemical vapour deposition
Linyang, JinkoSolar and SPIC. All of them use low- (APCVD), plasma-enhanced chemical vapour
pressure chemical vapour deposition (LPCVD) of deposition (PECVD) and physical vapour deposition
the poly-Si layer, a choice that is motivated by the (PVD) – have been developed, which allow single-
excellent passivation quality of LPCVD-deposited sided deposition of poly-Si. Furthermore, such
poly-Si layers, as well as by the availability of alternative technologies allow the etching of
corresponding deposition tools. However, it is the residual poly-Si deposition from transport belts,
authors’ understanding that significant challenges holders or carriers, either continuously during the
still remain, including the complex process process or entirely outside of the reactor. These
sequence as a consequence of the conformal techniques will be discussed further later on. ISC
poly-Si deposition in the LPCVD process, and Konstanz is evaluating many of these techniques
the necessity to deposit thick poly-Si layers for in order to develop a simple and high-throughput
achieving sufficient passivation in combination process that can be transferred to the market in the
with screen-printed firing-through metallization. coming years.
The latter procedure further reduces the service life
of the quartz tubes, one of the drawbacks of LPCVD Achieving higher voltages as in PERC
deposition of poly-Si, which will be discussed in A simple way of comparing different c-Si cell
more detail below. These challenges will need to technologies is to look at their maximum voltages
be overcome in order for TOPCon to compete with instead of efficiencies. The open-circuit voltage Voc
PERC, which requires not only achieving high is a reliable measure of the recombination at high
efficiencies but also fulfilling the specifications in carrier concentrations, and defines the upper limits
terms of throughput and yield. Typical specifications for both fill factor and cell power. A comparison
associated with these quantities for n-type solar cell of the measured efficiency, on the other hand,
production are shown in Table 1. is sometimes misleading, as several institutes
For a classical diffused nPERT cell, an efficiency
of 23% was unattainable. Therefore, all nPERT
“Standard PERC cell efficiencies are expected to
producers have switched (or are currently
switching) to TOPCon, where 23% is possible, reach their limits soon.”

78 www.pv-tech.org
 Passivated contacts | Cell Processing

Figure 4. Highest efficiencies, measured at ISFH, for large-area PERT (‘TOPCon’) and PERC solar cells in China in 2019.

Typical voltages for PERC solar cells are around

Efficiency [%] >23 680mV, with slight increases being possible when
using selective emitter technologies. A voltage
Throughput [wafers/sec] ~1.3
of around 700mV can be reached when applying
Capacity [MW/line] >200 passivated contacts to one polarity of the cell – for
Yield [%] >98 example by replacing a diffused rear side with
a silicon oxide poly-Si stack. The latter is much
Cost of cell (relative to PERC) <1.2
easier for n-type technologies, since in this case the
CAPEX (relative to PERC) <1.3 PCOl3 diffusion is on the rear side. Instead of using
Fig. 5 Size M4, M6 passivated-contact technology, a reduction of the
metallized surface area, for example by the use of
Table 1. Current manufacturing specification requirements for n-type cells [9]. point contacts, enables the voltage of the cell to be
increased. With a flat boron-diffused surface (rear
emitter PERT or IBC), it is possible to obtain voltages
of the order of 700mV. Such high voltages without
Two passivating contacts passivated contacts have been achieved by imec in
(also HJT) collaboration with Jolywood with their rear emitter
Advanced screen printing
nPERT [10], by ISC Konstanz with MoSoN cell
or one passivating contact
technology [11,12], and by ISC in collaboration with
SPIC with ZEBRA technology [13].
In order to reach 720mV, passivated-contact
technology is required for both polarities – as also
employed for heterojunction (HJT) cells. However,
PERC as of now, the equipment for such processes is very
expensive, so typical manufacturing specification
720mV+ demands cannot be fulfilled yet. To reach even
700mV+ higher efficiencies, new devices – such as four-
terminal tandem technologies with c-Si bottom solar
680mV+ cells and Perovskite solar cells on top – could be the
answer within the next five to ten years.
NOW 3-5y 10y
‘TOPCon’ (passivated-contact
technology) – what is it exactly?
Poly-Si passivated carrier selective contacts employ
Figure 5. Classification of solar cell technologies in terms of Voc. a thin silicon oxide layer to separate a highly
doped poly-Si layer from the bulk absorber of the
“The costs for TOPCon cells must not be more than solar cell, as shown schematically in Fig. 6. This
structure allows a high majority-carrier current
1.2 times the costs for PERC.” towards the metal electrode contacting the poly-
Si layer. The minority-carrier current, in contrast,
and companies report results from busbar-less is effectively blocked at the interface oxide, which
measurements, or they report active-area efficiencies suppresses interface state-mediated charge-carrier
by subtracting the shading of the front-side recombination at the metal contact.
metallization. Fig. 5 shows the three categories with The basic idea of using the passivated-contact
Voc ranges that can be achieved using low-cost solar concept (which is well known in bipolar transistor
cell processes today and that could be achieved in technology) for solar cells dates back to the 1980s.
the future. However, it did not receive much attention at that

Photovoltaics International 79
 Cell Processing | Passivated contacts
Fig. 6 and Table 2 (use Word doc table)

(a) (b)

n+ e h poly‐Si (n+)
SiOx
n recombination
c‐Si

c‐Si (n)
e h

Conventional diffused BSF c‐Si/SiOx/poly‐Si passivating contact

Efficiency
Figure 6. Schematic Cell type
cross sections (not to scale) of: (a) a conventional diffused back-surface field (BSF);Research Institute
(b) a layer stack / Company
for a poly-Si passivated
contact.
Tunnel layer passivated interdigitated back
25.2% contact (IBC) solar cell SunPower
Efficiency Cell type Research institute/company

25.2% Front-
Tunnel layer and IBC
passivated rear-contacted
solar cell tunnel SunPower

25.7%
25.7% layer passivated contact solar
Front- and rear-contacted TOPCon solar cell
cell FhG ISE
FhG ISE
(TOPCon)
26.1% Poly-Si on oxide (POLO) passivated-contact IBC solar cell ISFH
Polysilicon on oxide (POLO) passivated
Table 2. Highest efficiencies
26.1% achieved using the different
contact technologies.
interdigitated back contact (IBC) ISFH
solar cell
PERC 2010 Substrate Passivation AlOx Metallization Other
mc-Si AlOx Spatial ALD Evaporated Cell structure
Cz-Si Thermal SiO2 Batch ALD, Screen printed Process flow
Low O SiCx Remote LFC Stabilization
LID PECVD Plate Laser opening PID
PECVD Chemical opening Bifacial
Yield
Cost
TOPCon 2018 Substrate Thin oxide Poly-Si Metallization Other
p-type wet chemical LPCVD Evaporation Cell structure
n-type UV PECVD Screen printing Process flow
Thermal APCVD TCO UV degradation
Sputter PVD Plating PID
HWCVD Al paste Bifacial
EB PVD Yield
Cost

ALD = atomic layer deposition; LFC = laser-fired contact; LID = light-induced degradation; PID = potential-induced degradation; HWCVD = hot-wire chemical vapour
deposition; EB PVD = electron beam physical vapour deposition; TCO = transparent conductive oxide

Table 3. Open questions during PERC development starting in 2010, and during TOPCon development from 2018 onwards (adapted from Chen [17]).

time, as silicon solar cell efficiency was still more picked up the topic and contributed valuable
severely limited by other components of the device. insights in areas such as transport through the oxide
It is believed that SunPower was the first company (ISFH), the influence of doping profile variations
to commercially use passivated-contact technology (ECN), and alternative deposition methods (ANU,
to increase the efficiency of their IBC cell, but the SERIS, ISC, FZJ). Record efficiencies for small-area
company never disclosed any details on the cell laboratory cells were achieved by Fraunhofer ISE
architecture. with 25.7% [15] for two-sided solar cells, and 26.1% for
Passivated contacts for solar cells have been an IBC by ISFH [16] (see Table 2).
receiving ever-increasing attention since 2014, when
the group of Prof. Stefan W. Glunz at Fraunhofer How mature are passivated-contact
ISE researched the fundamental properties technologies in production?
of the passivation layer stack using modern Solar cell architectures which employ passivated
characterization methods [14]. Fraunhofer ISE also contacts (TOPCon) and are produced by different
coined the name TOPCon as the abbreviation for manufacturers today are very similar, even though
‘tunnel oxide passivated contact’ – initially for
a PECVD-deposited Si-layer stack with carbon
admixture, but eventually for all kinds of poly-Si
“For passivated-contact nPERT technology, the most
layer on a SiO barrier layer. Other research institutes critical process is the formation of the poly-Si layer.”

80 www.pv-tech.org
 Passivated contacts | Cell Processing
Fig. 7

ISC‐LPCVD Jolywood /SPIC TRINA Linyang BSG Linyang AlOx

sde & texture sde & texture sde & texture sde sde

BBr3 diffusion BBr3 diffusion BBr3 diffusion Clean & oxide Clean & oxide

Single-side etch Single-side etch Single-side etch LPCVD poly‐Si LPCVD poly‐Si

Clean & oxide Clean & oxide Clean & oxide POCl3 diffusion POCl3 diffusion

LPCVD n+ poly‐Si LPCVD poly‐Si LPCVD poly‐Si Wet process Wet process

Annealing P implant POCl3 diffusion SiNx rear SiNx rear

SiNx rear Anneal Anneal Texture Texture

Wrap‐around etch Wrap‐around etch Single-side etch BBr3 diffusion/BSG BBr3 diffusion
and cleaning

SiNx front AlOx front ALD SiNx rear Wet process Wet process

Metallization & FF Front & rear SiNx SiNx front SiNx front AlOx/SiNx front

Metallization & FF Metallization & FF Metallization & FF Metallization & FF

Figure 7. Different process flows for nPERT cells with a rear-side passivated contact.

choices had to be (and still have to be today)

Company Capacity [MW] carefully evaluated by considering the associated
Jolywood 2,200 cell efficiency potential and the matureness and
cost of the process. Diffusions of the front side
Yingli 800
are carried out by quartz tube diffusion, in a POCl3
JinkoSolar 800 atmosphere for PERC and in a BBr3 atmosphere
LG 500 for TOPCon. Other processes had to be (and still
have to be) evaluated too – such as the selection of
Linyang 500
substrate, passivation layers, stabilization treatments,
Trina 500 and opening of dielectrics, as well as metallization
SPIC 400 technology and respective pastes.
For passivated-contact nPERT technology, the
Total 5,700
most critical process is the formation of the poly-Si
Table 4. Companies with passivated-contact nPERT capacity in 2019 [9]. layer. Here the question remains how to achieve a
sufficiently doped one-sided poly-Si layer so as to
keep the process simple and cost effective. However,
Boron diffusion Poly-Si LPCVD the metallization remains challenging too, as lower-
Centrotherm Tempress temperature Ag pastes need to develop in order to
achieve a good contact without penetrating the thin
Tempress Semco
oxide layer. The company Toyal is even suggesting
Semco Centrotherm the use of low-temperature Al pastes to make this
Laplace Laplace process more cost effective.
Even though these challenges have not yet been
7-Star Polar
completely surmounted, several companies are
P&Tech P&Tech already moving towards production; see Table 4 for
a summary. Jolywood has the highest production
Table 5. Two key steps for TOPCon and selected equipment suppliers [9].
capacity; however, this relies on fairly complex ex
situ doping of the LPCVD poly-Si layers by ion
the processing sequences may differ. Table 3 depicts implantation. The total capacity of all passivated-
a collection of different technology options that contact nPERT producers is about 6GW, of which a
companies have been able to choose from during total of 4GW is accounted for in 2019.
PERC development since 2010, and now during As already discussed, the COO falls short of being
TOPCon development since about 2018 [17]. These competitive with PERC, as all the process flows

Photovoltaics International 81
 Cell Processing | Passivated contacts

involved are still too complex. Fig. 7 shows several “Doping of the poly-Si layer can also be achieved via
process flows published by the major TOPCon
producers, along with ISC Konstanz’s LPCVD
in situ doping during deposition, but at the expense
reference process. of a reduced deposition rate.”
What the different processes have in common is
that the POCl3 diffusion is replaced by the formation has published an average efficiency of 23.0% with
of a thin interface oxide and a subsequent LPCVD an average Voc of 702mV [18]. The record TOPCon
poly-Si deposition. The resulting silicon thin film efficiency from Trina of 24.58% with a voltage of
is then doped ex situ via POCl3 diffusion (Trina, 717mV is suspected to have been achieved using a
Linyang) or via ion implantation ( Jolywood and selective passivated contact on the front as well,
SPIC). As illustrated by ISC Konstanz’s process flow, similarly to the record ‘PERC’ cells from LONGi.
the doping of the poly-Si layer can also be achieved A very good summary for all high-efficiency solar
via in situ doping during deposition, but at the cell technologies on the market is also given in the
expense of a reduced deposition rate. report from Tayiang News [19].
A significant difference between the process flows In order to obtain not only high efficiencies but
of Linyang and the other companies is the sequential also low COO, several bottlenecks of the TOPCon
arrangement of the emitter diffusion and the poly- process routes described above need to be addressed.
Si deposition. Whereas Linyang uses a process flow One of the most important questions concerns the
in which the BBr3-diffusion is performed after the technology for poly-Si deposition, as detailed in the
LPCVD deposition of poly-Si, the other published following section.
processes implement the poly-Si deposition after
the BBr3 diffusion. Both approaches have different Comparison of poly-Si deposition
advantages and challenges concerning the front- and technologies
rear-side passivation. As of now, most cell manufacturers are focusing
Emitter passivation can be achieved using either on developing passivated contacts using LPCVD
a stack of AlOx/SiNx or a combination of boron deposition of poly-Si (Table 5). This choice is
silicate glass (BSG) and SiNx, as exemplified by the motivated by both the excellent passivation quality
two different Linyang process routes. Even though achieved with these layers, and the availability
record efficiencies of above 24% can be realized by of industrial-scale deposition tools developed for
an adapted TOPCon process, average efficiencies in the deposition of poly-Si in the semiconductor
production vary between 22.8% (SPIC/Voc = 695mV) chip industry. While the results based on LPCVD
Fig. 8
and 23.3% (Linyang with AlOx/Voc = 695mV). deposition of poly-Si published by various cell
Jolywood’s average efficiency in production is manufacturers and research institutes are promising,
reported to be 23.1% with a Voc of 700mV, while Trina this technology presents several challenges.

Deposition technique Features Schematic

‐ Hot wall CVD

‐ No ion impact needed
‐ Very conformal
LPCVD
deposition
BUT: both sided and tubes
cracking

‐Single-sided deposition
PECVD ‐Lower temperature
than LPCVD
BUT: wrap around

‐ Much higher deposition

rate than PECVD
‐Single sided
HWCVD BUT: machine not
industrial

Figure 8. A selection of process technologies for the creation of poly-Si layers for passivated-contact technology.

82 www.pv-tech.org
 Passivated contacts | Cell Processing

An inherent disadvantage with many cell concepts situ boron-doped layers have been successfully
is the conformal deposition of the layer, which integrated into p-type TOPCon cells with sputtered
requires dedicated process steps for single-side poly-Si and full-area metallization on the rear side,
etching of poly-Si. Moreover, the lifetime of the achieving 23% efficiency [24]. Excellent surface
quartz tubes is usually limited to a total deposition passivation has also been demonstrated for ex situ
of around 100µm of poly-Si, owing to the increasing phosphorus-doped layers produced by sputtering of
internal stress, which eventually leads to tube intrinsic silicon [25]. Sputtering of silicon could be
breakage. Since typical poly-Si layer thicknesses an attractive alternative, offering a high-throughput
of 150–300nm are currently required for sufficient process that produces a hydrogen-free silicon layer,
passivation after metallization [20,21], the quartz without the need for any toxic gases and hence
tubes must be replaced after ~700 runs. avoiding blistering of the layers.
In addition, the wafer throughput is limited by
the deposition rate of around 3–6nm/min, which is Summary and outlook
further reduced when adding phosphine for in situ Passivated-contact technology implemented in
phosphorus doping of the layer. The latter issue can nPERT structures, often referred to as TOPCon
be mitigated by employing ex situ doping, e.g. using in China, is likely to be the next step after PERC
POCl3 diffusion after the poly-Si deposition. technology, which is slowly approaching its
Given the challenges concerning the conformal limits. The total production capacity of TOPCon
deposition and the throughput of the LPCVD in 2019 was 5.7GW. However, there are still many
process, several alternative technologies have been challenges remaining to make this promising
investigated by research institutes around the world technology cost effective and competitive with
(Fig. 8). Possible single-sided deposition methods for PERC. The process needs to be simplified, mainly
poly-Si include different chemical vapour deposition by developing high-throughput processes for
processes (PECVD, HWCVD and APCVD), and single-sided deposition of poly-Si with in situ
various physical vapour deposition (PVD) processes, doping. Some progress in silver paste composition
e.g. sputtering of silicon. Industrial tools are available also has to be made in order to further reduce the
for all of these processes, but not necessarily for the poly-Si layer thickness while maintaining excellent
specific application of depositing highly doped and passivation after metallization. On top of that,
ultrapure silicon thin films. Research on optimizing replacing Ag metallization either partly or fully
these alternative processes for the production of by Al metallization, without cannibalizing cell
solar cells with passivated contacts is therefore performance, as suggested (for example) by Toyal,
currently gaining increasing attention. would lead to a further essential cost reduction.
Silicon layers deposited via PECVD have been PV technologies nowadays are moving towards
successfully integrated into TOPCon nPERT cells, higher front-side power and higher bifaciality,
leading to efficiencies of around 23% [22]. One in order to save balance of system (BOS) costs in
of the main challenges of the PECVD-deposited utility-scale systems. Since a higher voltage results
silicon layer is the incorporation of hydrogen, which in a lower temperature coefficient, solar cells with
can lead to blistering of the layer. To avoid such passivated contacts, and other high-voltage devices
effects, the deposition conditions – in particular with high bifaciality factors, will become important
temperature, gas flows, pressure and plasma power in the future in order to achieve
– need to be carefully optimized, making the 1US¢/kWh in horizontal single-axis tracking
optimization of the process more challenging than (HSAT) bifacial systems. Such a low levelized cost of
LPCVD depositions. Additionally, the wrap-around electricity will allow PV to enter the sustainable TW
has to be minimized by dedicated carrier designs. era in the coming years.
HWCVD deposition of the silicon layer uses
hot wires to dissociate precursor gases. As with
PECVD, in situ doping of the layers can be achieved,
for example using phosphine and diborane as References
the dopant source. One of the key advantages of [1] PV ModuleTech Conf. 2020, Kuala Lumpur, Malaysia.
HWCVD is the potential to deposit a poly-Si layer [2] Hermle, M. 2017, “Silicon solar cells – Current
with excellent surface passivation at very high production and future concepts”, Proc. Ann. Conf.
deposition rates of up to 42nm/min [23]. Euro. Tech. Innov. Plat. Photovolt. – PV Manufac. Euro.,
Finally, PVD technologies, such as sputtering Brussels, Belgium.
of silicon, can also be used as an alternative. In [3] PV InfoLink [https://en.pvinfolink.com/].
[4] Stodolny, M.K. et al. 2018, “Novel schemes of p+
poly-Si hydrogenation implemented in industrial
“Solar cells with passivated contacts, and other high- 6” bifacial front-and-rear passivating contacts
voltage devices with high bifaciality factors, will solar”, Proc. 35th EU PVSEC, Brussels, Belgium
[10.4229/35thEUPVSEC20182018-2CO.10.3].
become important in the future in order to achieve [5] Ingenito, A. et al. 2018, “A passivating contact
1US¢/kWh in HSAT bifacial systems.” for silicon solar cells formed during a single firing

Photovoltaics International 83
 Cell Processing | Passivated contacts

thermal annealing”, Nat. Energy, Vol. 3, pp. 800–808 10.1063/1.5089650].

[https://doi.org/10.1038/s41560-018-0239-4]. [24] Cuevas, A. et al. 2018, “Silicon solar cells by
[6] Bende, E.E. et al. 2019, “P-IBC: Combining PERC ‘DESIJN’”, Proc. 35th EU PVSEC, Brussels, Belgium.
and TOPCon”, Proc. 9th Metall. Interconn. Worksh., [25] Hoß, J. et al. 2019, AIP Conf. Proc., Vol. 2147, 040007
Konstanz, Germany. [https://doi.org/10.1063/1.5123834].
[7] Haase, F. et al. 2019, “Transferring the record
p-type Si POLO-IBC cell technology towards an
industrial level”, Proc. 46th IEEE PVSC, Chicago, About the Authors
Illinois, USA. Dr. Radovan Kopecek obtained his
[8] 2019, Plenary session, 29th Int. PVSEC, Xi’an, Diploma in physics at the University
China [http://www.pvsec.org/_userdata/Final_ of Stuttgart in 1998. He also studied
report/Final_Report_PVSEC-29.pdf]. at Portland State University (Oregon,
[9] JinkoSolar 2019, Presentation, CSPV Conf., USA) and obtained a master’s there in
Shanghai, China. 1995. He finalized his Ph.D.
[10] imec/Jinkosolar 2019, Press Release [https:// dissertation in Konstanz in 2002, and was a group
www.imec-int.com/en/articles/imec-and-jolywood- leader at the University of Konstanz until the end of
achieve-a-record-of-23-2-percent-with-bifacial-n- 2006. Dr. Kopecek is a co-founder of ISC Konstanz,
pert-solar-cells]. where he has been working as a full-time manager
[11] Kopecek, R. et al. 2019, “Low-cost standard nPERT and researcher since 2007, and is currently the head
solar cells towards 23% efficiency and 700mV voltage of the Advanced Solar Cells department. He has also
using Al paste technology”, Photovoltaics International, been on the board of directors of EUREC since 2016.
42nd edn.
[12] Peng, Z.-W. et al. 2019, “Investigation on Dr. Jan Hoß studied physics at the
industrial screen-printed aluminum point contact Karlsruhe Institute of Technology
and its application in n-PERT rear junction solar (KIT), Germany. He received his Ph.D.
cells”, IEEE J. Photovolt. in 2017 from ETH Zurich, Switzerland,
[13] PV-Tech 2019, News Report (in Chinese) [https:// where he contributed to the
www.pv-tech.cn/news/Yellow_River_Hydropowers_ development of a semiconductor
IBC_products_rank_among_the_top_in_the_world]. detector for high-energy particles for the Compact
[14] Feldmann, F. et al. 2014, “Passivated rear contacts Muon Solenoid (CMS) experiment at CERN’s Large
for high-efficiency n-type Si solar cells providing Hadron Collider. In 2017 Dr. Hoß joined ISC Konstanz
high interface passivation quality and excellent as a research associate, where he is involved in
transport characteristics”, Sol. Energy Mater. Sol. Cells, developing n-type solar cells with charge-carrier
Vol. 120, pp. 270–274. selective contacts, with a major focus on alternative
[15] Richter, A. et al. 2017, “N-Type Si solar cells with poly-Si deposition technologies.
passivating electron contact: Identifying sources
for efficiency limitations by wafer thickness and Jan Lossen studied physics at the
resistivity variation”, Sol. Energy Mater. Sol. Cells, Vol. Universities of Freiburg and Cologne
173, pp. 96–105. in Germany. He graduated in 2003
[16] Haase, F. et al. 2018, “Laser contact openings for with a thesis on the hot-wire chemical
local poly-Si-metal contacts enabling 26.1%-efficient vapour deposition of microcrystalline
POLO-IBC solar cells”, Sol. Energy Mater. Sol. Cells, Vol. silicon layers. Over a period of more
186, pp. 184–193. than ten years, he worked in different capacities on
[17] Chen, Y. 2019, Presentation, PV CellTech, Penang, the production and development of crystalline silicon
Malaysia. solar cells at ErSol/Bosch Solar Energy AG. Since 2014
[18] Chen, Y. et al. 2019, “Mass production of Jan has been a senior scientist and project manager
industrial tunnel oxide passivated contacts for process transfers at ISC Konstanz, where he is
(i‐TOPCon) silicon solar cells with average head of the Passivated Contacts group.
efficiency over 23% and modules over 345 W”, Prog.
Photovolt: Res. Appl., Vol. 27, pp. 827– 834 [https://doi.
org/10.1002/pip.3180].
[19] Tayiang News 2019, News Report [http:// Enquiries
taiyangnews.info/reports/high-efficiency-solar- Dr. Radovan Kopecek
cells-2019/]. International Solar Energy Research Center
[20] Chaudhary, A. 2019, AIP Conf. Proc., Vol. 2147, Konstanz e.V.
040002 [https://doi.org/10.1063/1.5123829]. Rudolf-Diesel-Straße 15
[21] Chen, D. et al. 2020, Sol. Energy Mater. Sol. Cells, Vol. D-78467 Konstanz, Germany
206, 110258 [doi.org/10.1016/j.solmat.2019.110258].
[22] Nandakumar, N. et al. 2019, Prog. Photovolt: Res. Tel: +49 (0) 7531 / 36 183 – 22
Appl., Vol. 27, No. 2, pp. 107–112. Email: [email protected]
[23] Li, S. et al. 2019, Appl. Phys. Lett., Vol. 114 [doi: Website: www.isc-konstanz.de

84 www.pv-tech.org