精华版(总第6期)

新品发布 AEG公司Thyrobox M能源解决方案

|

欧瑞康DRYVAC真空泵

|

蓝星有机硅CAF 530系列封装产品

Photovoltaics International 2010年总第6期

晶体硅太阳能电池发射极 形成的若干方法 ECN太阳能能源部介绍该类技术

多晶硅硅片制造流程上的 降本增产策略 PV-Tech博客

对全球上网电价补贴调整趋势的调查研究

www.pv-tech.org

406565-Photovoltaics International advert AW.indd 1

07/09/2010 07:45

卷首语 出版机构： Solar Media Ltd. Trans-World House, 100 City Road London EC1Y 2BP, United Kingdom 电话：+44 207 871 0123 传真：+44 207 871 0101 电邮：info@pv-tech.org 网址：www.pv-tech.org 出版人： David Owen 责任编辑：Cathy Li 翻译编辑：Huangye Jiang, Liqiong Hu, Yang Li, Chuanxi Yang 版面设计：Daniel Brown 制作经理：Tina Davidian 英文编辑：Síle Mc Mahon, Tom Cheyney Mark Osborne, Emma Hughes 亚洲销售代表：James Park 欧美销售代表：Adam Morrison, Neill Wightman, Graham Davie, Daniel Ryder, Gary Kakoullis, David Evans, Nick Richardson 市场经理：Joy-Fleur Brettschneider 免责及限责条款： 虽然我们已尽最大努力来提高Photovoltaics International的质量和准确性，我们的作者也以 极度严谨的态度来对待每一篇文章，我们对所 有原样提供的内容均不做担保。Photovoltaics International对其中包含的第三方内容不负有 责任；对广告内容中的错误、删节或不精确之 处不负有责任，对在杂志中因广告需要出现 的互联网址的可用性不负有责任。本杂志提 供的数据和信息仅供参考用途，Photovoltaics International杂志、其附属机构、信息提供者 及内容提供者均不对任何基于本杂志信息作出 的投资决策或者由本杂志信息所得出的结论 负责。

目前，全球太阳能市场的需求量正经历着大幅上涨，同时，供应链上各主要光伏 生产商的产品也都已销售一空，甚至是那些已完成大规模产能扩张的企业产品也所剩无 几，其他生产商旗下工厂的开工率也极高。欧洲各终端市场仍旧是产业增长的主要动 力，特别是德国、意大利、法国和捷克共和国等。 同时，亚洲各光伏供应商也正享受着2010年全球市场蓬勃发展所带来的果实。据市 场调研公司Solarbuzz近期公布的第二季度数据来看，六家中国大陆企业的太阳能电池 出货量占到了全球总量的55%，同比去年的43%有所上涨。 这些企业的关键性发展还包括在保障进一步降低生产成本能力并改善晶硅电池总体 效率的同时，进行大幅增产，其发展速度要远比其西方同业者快得多。 在本期杂志所呈现的文章中，重点讨论了亚洲光伏制造商是如何借助市场供应趋势 自我发展的。同时，关于低成本硅片技术及发展策略的讨论也着重强调了业内对导致硅 片业务大获成功的减本策略的关注。此外，在本期刊载的由ECN所提供的技术文章中， 作者详细地讨论了关于太阳能电池选择性发射极的话题，该技术目前在领先大规模电池 制造商(多为亚洲企业)中较为流行，是一项生产工艺上的创新，可将转换率提高至近 20%。 随着各市场调研公司急于将2010年的新增安装量提至16GW以上，亚洲地区正迅速成 长为可满足全球市场需求的主要制造中枢。 马克·奥斯本（Mark Osborne） 高级新闻编辑 Photovoltaics International杂志/PV-Tech.org网站

封面图片：位于阿斯彭山滑雪度假村的光伏系 统，NREL友情提供 图片拍摄：阿斯彭滑雪公司 印刷公司：Shing-Hsiang Printing Co. Photovoltaics International精华版 2010年总第6期 ISSN 1757-1197

目录

使用限制： 本杂志受国际版权法规及商标法规保护。任 何人不得以任何形式修改、复制、衍生、重 印、发表、转载或散布本杂志内容的任何部 分。Photovoltaics International杂志社要求读 者遵守相关版权要求及所有权声明，仅使用 本杂志作为个人的非商业用途。如读者需要 使用本杂志作为非个人、非商业用途，须经 Photovoltaics International公司的书面许可。

2 国际光伏新闻 12 新品发布

19 ECN: 晶体硅太阳能电池发射极形成 的若干方法

感谢Sputnik和 Solarmax公司 友情提供图片

图片版权归属： Sonel d.o.o.

14 Photovoltaics International: 多晶硅硅片制造流程上的降本 增产策略

Photovoltaics International中文精华版特 为PV Taiwan 2010太阳光电展量身定做， 在保留杂志完整版风格的基础上，本着深 入浅出、精益求精的原则，刊载了部分技 术文章、产品简介及舆论导向。本刊完整 版为光伏领域门户期刊，每季一期，期期 精彩，不容错过。 全年订阅价：199美元 详情垂询亚洲销售代表：James Park 电邮：jpark@pv-tech.org

10 杜邦 (DuPont) 专家观点

23 PV-Tech博客：对全球上网电 价补贴调整趋势的调查研究 Emma Hughes

Photov ol tai c s I nter nati on al

1

国际光伏新闻

—— 阅读更多每日光伏新闻，请登录: cn.pv-tech.org

世界首次：SoloPower柔性CIGS组件获得UL认证 柔性太阳能光伏技术近日获得了一项突破性进展，SoloPower公司 的轻质柔性铜铟镓硒(CIGS)组件获得了美国保险商实验室 (Underwriter Laboratories Inc)的认证，这是薄膜行业首次取得这样的成果。这也是首批 进入欧洲和北美市场的大功率柔性组件中首个获得UL认证的产品。 UL 1703标准认证是一项衡量光伏组件安全生产的标准，柔性薄膜组件 将在一处独立实验室中接受UL 1703标准测试，产品在经历一系列严苛的测试 之后才会被授予UL认证。SoloPower还表示其进行过大量的内部测试，在安全 性、质量和可靠性方面都比UL测试的标准更加严格。 “SoloPower柔性CIGS组件的认证是实现轻质、大功率的柔性太阳能组件 大规模应用的重要一步，这种组件在扩大屋顶太阳能市场和降低系统成本方 面有着巨大潜力。这是行业一个重要的里程碑。在美国国家可再生能源实验 室(NREL)研究了30年CIGS薄膜光伏组件之后，我很欣慰能看到这项技术终于 走 向了成熟。”NREL的首席科学家隆美尔·诺菲(Rommel Noufi)博士表示。 SoloPower旗下的柔性光伏组件包括SFX1型组件(80Wp，0.3m x 2.9m， 2.3kg)，SFX2型组件(170Wp，0.3m x 5.8m，3.6kg)，和SFX3型组件 (260Wp，0.9m x 2.9m，6kg)。

薄膜制造

杜邦PV8600系列产品将被用 于夏普太阳能电池组件 杜邦光伏解决方案业务部门 (DuPont Photovoltaic Solutions)日 前公布，其近期研制的离聚物密封板 (ionomer encapsulant sheet)将被夏 普公司用于其薄膜太阳能电池组件， 这一在日本名为“Himilan ES”的产 品在其他国家和地区则被称为PV8600 系列。 “基于离聚物材料从根本上来 说，具有更强的抗潮、防电流泄漏和 防褪色性能。”杜邦Encapsulants全 球业务总监的史蒂夫•克拉夫(Steve Cluff)表示，“它们已被十分成熟 地用于玻璃-玻璃组件设计，但用于 粘合基于聚合物的背板时，例如杜邦 Tedlar和PET系列产品，则需要一定改 进，我们在新款杜邦PV8600系列面板 中便采用了这种改进。”

杜邦表示，此密封材料将在今年 晚些时候向日本以外的地区开放顾客 评价。

Sunvalley Solar拟采用天威 太阳能非晶硅薄膜组件 美国光伏安装商Sunvalley Solar 公司近日将保定天威太阳能公司的组 件加入到其供货目录中。Sunvalley Solar此前已与阿斯特太阳能和中电 电气签有多晶和单晶组件供应协议。 天威太阳能将向Sunvalley Solar提供 数量不详的非晶硅薄膜组件，该产品 已经通过加州能源委员会(California Energy Commission)的认证。

天威太阳能首席技术官麦耀华博士代 表公司为客户介绍生产工艺

台积电拟销售自产铜铟镓硒 薄膜组件，首座200MW工厂 已动工开建 采用了杜邦光伏解决方案的PV8600系 列离聚物密封板

2

w w w. pv -te c h .o rg

随着公司业务逐步从此前在半导 体工业内的单一型硅片代加工业务中 分离出来，台湾积体电路股份有限公

安装SoloPower柔性CIGS组件

司(TSMC)已于今日(九月十六日)破土 动工进行其首座200MW工厂的建设，并 计划将所产铜铟镓硒(CIGS)薄膜组件 销往全球市场。该项目初始投资金额 为2.58亿美元，用于建造薄膜太阳能 电池研发中心和启用Stion公司技术的 制造工厂。作为一家生产薄膜产品的 新兴企业，Stion公司已于台积电达成 协议，并将成为其制造及技术发展的 合作伙伴。 台积电还宣布，计划为工厂进 行二期建设，将位于台中的中部科学 工业园区内的铜铟镓硒产品产量扩至 700MW以上，同时增加约两千名员工。 该园区是公司前沿半导体，以及近日 宣布的进军发光二极管(LED)市场计划 的中心。 据称一期的设备移入工作计划将 于2011年第二季度完成，并将在2012 年达到200MW的初始年均产量。公司并 未披露规模更大的二期产能扩张计划 的具体时间表。 “台湾积体电路公司始终致力于 成为技术先锋，我们的太阳能业务也 不例外。”台积电新事业董事长蔡力 行(Rick Tsai)博士表示，“在此研发 中心进行的研究，将有助于我们实现 提供领先薄膜电池解决方案的目标。 并且凭借台积电在制造实践技术方面 的经验，此工厂的生产将为我们铺平 道路，以成为太阳能光伏组件领域的 顶级供应商。” 据了解，首套铜铟镓硒设施的占 地面积将达110,000平方米，总生产面 积可达78,000平方米。 台积电的主席兼首席执行官张 忠谋(Morris Chang)表示：“此太 阳能研发中心及工厂的建造，以及我 公司在其附近的Fab 15超大型晶圆厂

ASYS SOLAR presents: New performance enhancing technologies SEI 02

Laser edge isolation system with improved throughput

SLD 01

Laser drilling of vias for MWT and EWT

SLS 01

Backside contact isolation for MWT

SAS 01

Alignment system for printing SE or MWT on XS1 or XS2

Si-Ink Print & Dry Island for Innovalight‘s proprietary SE Cougar Process

台湾积体电路股份有限公司总部

(Gigafab)意味着中部工业园区将成为 台积电大部分先进及创新型生产的基 地。” 太阳能装置

SunEdison与韩国省份签署 400MW太阳能电站开发谅解 备忘录 美国MEMC电子材料公司的子公司 SunEdison与韩国东南部省份庆尚南道 政府签署了在其辖区建设400MW光伏发 电设施的谅解备忘录。这份非约束性 的谅解备忘录还需要双方进一步协商 并达成最终协议，协议计划在公共土 地和建筑屋顶安装光伏发电系统。在 这项潜在的交易面前，公司之前的发 展计划显得相形见绌。 庆尚南道政府表示其将在选择合 适的土地和建筑，以及项目授权和审 批过程方面向SunEdison公司提供支 持。整个项目预计在2013年底完成。 交易的财务细节并未被披露。 SunEdison公司将拿出项目的部分 利润用于当地经济的发展，公司还鼓 励庆尚南道的其他企业加入到太阳能 电站的安装和建设中来。 在7月底MEMC最新一季财报发布的 电话会议上，母公司表示SunEdison在 建的111.5MW项目将于2010年底完工。 如果这项与韩国的交易获得批 准，其带来的项目总量将大约相当于 SunEdison年终投资预期的两倍。这也 将是东亚地区最大的光伏系统安装一 揽子协议之一。

“能收到CE和VDE对于DGM-S460 系列的认证，我们感到十分激动。” Direct Grid公司董事长弗兰克•库珀 (Frank Cooper)表示，“此举在业内 尚属首例，同时也使得公司成功已跻 身于欧洲公共电网部署的500W级别之 中。顾客也可充分认识到分布式逆变 器由于无需配备组件电源而具有的结 构优势，同时，深入理解此类产品通 常所带有的在可靠性上的限制。” 库珀先生还表示：“鉴于此次 认证的获得是公司‘进军市场’计划 中所取得的重要里程碑，公司正策划 举办一场全球范围的产品发布会。从 而，在接下来的几周内，公司预计可 获得其他几份地区性标准认证，其中 包括美国市场上的UL认证等。”

宝威电源(深圳)将在菲尼克 斯建造1吉瓦逆变器生产工厂 宝威电源(深圳)有限公司日前选 择亚利桑那州菲尼克斯(Phoenix, AZ) 作为其新的生产设备站点。该厂将为 公司生产光伏及风能逆变器，包括单 相(2-6千瓦)和三相串型逆变器，以及 NEMA 3R 250千瓦，300千瓦和400千瓦 集中型逆变器。 公司表示，亚利桑那州设备的制 造将于十月份启动，截止到2011年中 旬，该站点的逆变器年产量将达到1吉 瓦。新工厂预计将为亚利桑那州，乃 至全国提供超过 350份工作。 宝威电源(深圳)表示，凭借加拿 大制造厂的启动，以及其在欧洲厂房 的扩建，公司全球产量正日益增加， 这将使得截止至年底，该公司逆变器 产量将超过4吉瓦。 该公司还计划，用专门为亚洲市 场设计的产品开启公司在中国第四季 度的生产。

Direct Grid公共事业等级太 阳能光伏微型逆变器获得CE 和VDE认证 Direct Grid Technologies公司 日前收到欧测国际认证(CE)和德国电 气工程师协会认证VDE0126-1-1对于 其DGM-S460系列产品的认证。该系列 公共事业等级、并网光伏微型逆变器 是为大型商用及太阳能场的设备而设 计。Direct Grid公司表示，此逆变器 将率先通过其法国的原始设备合作伙伴 One Network Energies公司在欧洲进行 分销，同时，公司的其他合作协议的商 讨也接近尾声，并将于近期公布。 4

w w w. pv -te c h .o rg

宝威电源出品的Aurora逆变器系统

SunEdison拟为亚利桑那州 建造14.5兆瓦电厂 SunEdison公司通过不断扩展一 系列大规模安装项目，于近日赢得了 一份电站合同。合同表示，该公司将 为坐落于亚利桑那州图森市(Tucson)

在戴维斯-蒙森空军基举办的空展

的戴维斯-蒙森空军基地(DMAFB)安装 14.5兆瓦的落地式太阳能装置。根据 协议规定，SunEdison将负责此52.6 万平方米(约130英亩)项目的财政、 设计、建造、运营和维护工作。该电 站将为戴维斯-蒙森空军基地提供多达 35%的能量需求。 SunEdison还与戴维斯-蒙森 空军基地缔结一项长期土地租契协 议，SunEdison将有权使用该基地的土 地，作为回报SunEdison将以既定价格 为基地供应电力。 “凭借我们强劲的资金实 力，以及过去适时的大规模部署业 绩，SunEdison公司能提供合理的 太阳能解决方案，而不需要前期 投资费用以及长期可预见的能源价 格。”SunEdison销售部副总裁史密斯 先生(Jaime A. Smith)表示。“能被 授予这个合同，SunEdison感到十分自 豪，并且期望能与戴维斯-蒙森空军基 的共同合作完成太阳能电厂的建设并 启动该电厂。”

十亿瓦规模：加州能源委员 会批准Solar Millennium建 造1吉瓦Blythe CSP电站 日前，Solar Millennium公 司提出的在里弗赛德县(Riverside County)建造并运营Blythe太阳能项 目的申请获加利福尼亚州能源委员会 (CEC)一致通过。建成后，此1GW聚光 型太阳能(CSP)电厂将成为世界上最大 的太阳能发电站。整个电厂计划由四 个250兆瓦抛物线型槽的聚光太阳能部 件组成，其中前两部分的建造计划将 于今年年底启动。 尽管CEC的同意十分关键，公司还 须获得一项“决策记录”，以证明其 获得了由联邦土地管理局批准此项目 的“准许通行权”，该证明预计将于 今年秋季获批。 Solar Trust of America公司的 项目发展机构表示，该公司还在为首 期500兆瓦的项目与美国能源贷款担保 项目部(U.S. Department of Energy Loan Guarantee Program)积极协调， 已顺利完成融资工作。 公司表示，该项目在建造期间将 创造约2500个工作岗位，一旦1GW的设 施全部完工，将创造超过200个永久性 工作岗位。项目完成后, 此座干燥冷 却式聚光型太阳能电厂的年产量将足 可满足超过三百万户家庭供电需求。

DuPontTedlar(TWN _Final)0804-2 copy.pdf 1 04/08/2010 09:39:05

C

M

Y

CM

MY

CY

CMY

K

。

Tedlar®

Solar Millennium’s Andasol 1，欧 洲第一家抛物槽聚焦式太阳能发电站

Solar Trust公司表示，其在加州 和内华达州境内共有九处已处于发展 后期的公共事业规模太阳热能项目。 Blythe电厂是最近获得CEC批准 的第三个主要的聚光型太阳能场。之 前两座分别是NextEra Energy公司提 议的250MW“灯塔”(Beacon)计划和 Abengoa的250MW莫哈维(Mojave)太阳 能计划。在今年年底之前，委员会还 将对另外几项聚光型太阳能项目进行 审批。

说比较陌生的材料，比如导电胶和导 电背板。我们很高兴我们的MWT组件技 术获得了南德意志集团的认证，我们 相信再过不久就会有使用这项技术的 组件问世。” “自从我们在2008年开始推销 我们的设备以来，市场一直对这项 独特的技术抱有极大的兴趣”，荷 兰设备制造商Eurotron的销售总监 Bram Verschoor表示，“MWT组件通过 认证之后我们就能够销售一体化生产 线。Eurotron公司的生产线使用焊盘 组装MWT组件，每条生产线的年产能可 达150MWp，而使用串焊机的传统生产 线的产能只有20MWp。” IEC61215是一项业内公认的工业 标准，通过此标准的组件表明其经过 了一系列测试并可以在多种气候条件 下长期使用。

光伏组件

焦点订单：中国光伏集团拟 向中国境内太阳能项目提供 20MW组件 组件制造商CNPV(中国光伏集团) 将在2012年向临沂巨皇新能源科技发 展有限公司提供20MW的光伏组件，用 于中国山东省境内若干1MW项目的建造 上。集团将在2010年10月份开始首批 1MW组件的供应，并将在2011年和2012 年期间陆续完成余下19MW的供应。 “首个并网系统将全面启用公 司的优质太阳能组件系列产品进行 连接。”CNPV首席运营官、首席技 术官兼董事会成员乔德瑞先生(B. Veerraju Chaudary)表示，“尽管公 司现有的优质产品系列产量已占据了 市场的领先地位，此套定制系统可在 此基础上进一步将产量提高25%。对于 我们来说，能进一步参与到山东省当 地的太阳能发电工作中来，是一件令 人十分高兴的事情。”

ECN金属缠绕式组件通过IEC 61215认证 荷兰能源研究中心(ECN)的金属 缠绕式(MWT)组件技术近日通过了IEC 61215工业标准认证。作为联合研究框 架协议的一部分，该项认证同时获得 了另一家未公开身份的光伏企业的认 可。 ECN太阳能部门的主管保罗·迈尔 斯(Paul Wyers)博士表示，“经过多 年的研发努力后，我们对自身的背电 极组件的性能和坚固性充满信心。背 电极组件使用了目前最先进的生产技 术。我们使用了一些对晶体硅组件来 6

w w w. pv -te c h .o rg

TÜV莱茵标准验证机构印度班加罗尔 总部

此外，我实验室旗下180名专家所 拥有的专业技能可为全球所有的客户 提供系统、组件及零部件的测试与认 证。” TÜV日本实验室光伏产品测试主任 Stefan Kiehn表示：“TÜV旗下各地实 验室之间均有密切的合作，共同为全 球业务的发展添砖加瓦。对于印度实 验室的开设，我们感到十分高兴，此 举将有助于我们满足亚洲市场不断上 涨的需求并继续为客户提供客制化的 解决方案以满足其需求。” 工厂设施

丹福斯宣布诺堡总部扩张计划

转换效率为17.0%的ECN金属缠绕式 组件(MWT)

TÜV实验室于班加罗尔增设 太阳能产品测试实验室 TÜV莱茵标准验证机构近日正式 在印度班加罗尔电子城(Electronics City in the Indian city of Bangalore)开设其第七处太阳能组件 系统测试实验室。TÜV为此实验室投资 200万欧元，将为印度的新兴太阳能产 业提供相关服务。 此测试中心占地2000平方米，包 括500平方米的室外测试场地，并加装 有五处人工气候室以及两套太阳能模 拟器。凭借这些设备，该实验室已成 为南亚经济圈内最先进的光伏产品测 试中心。 TÜV实验室首席执行官弗里德 里希·埃克(Friedrich Hecker)表 示：“我实验室在太阳能产业内进行 投资项目的宗旨是为所有蓬勃发展的 市场提供便捷的服务，并满足各家公 司对于测试规模的需求。印度市场是 一个不容忽视的市场。

在经历了被公司称作十分强 劲的销售阶段之后，丹佛斯太阳能 (Danfoss Solar)日前宣布，计划通过 将业务转移至位于丹麦诺堡的公司总 部，以利用其更大规模的设施来扩大 产量及物流运输能力。丹佛斯公司表 示，其在丹麦格罗斯腾(Gråsten)和森 讷堡(Sønderborg)的供应链设施已无 法满足公司过大规模的业务。该公司 不仅计划在2011年将生产能力增加至 3.5吉瓦，还将拥有能力在数月内将装 配尺寸翻倍。 新厂址预计于2011第一季度投产 使用，而搬迁将在明年年底完成。另 外，丹佛斯还将扩建其位于中国和美 国各地的设施，以促进太阳能逆变器 的市场需求。 丹佛斯董事长托尔斯•皮特森 (Troels Petersen)表示:“扩大我 们在丹麦、美国及中国的业务活动， 有助于丹佛斯电力电子部门(Danfoss Power Electronics)实现其业务增长 和销售业绩的目标进而完成丹佛斯集 团(Danfoss Group)所制定的新战略目 标。”

Solarwatt组件生产能力翻倍 至400兆瓦：新生产线组件 单位生产时间仅为28秒 位于德累斯顿(Dresden)的 Solarwatt公司日前，正式启动了被誉 为“世界最现先进组件生产设备”的

PV-tech.org 每日光伏新闻

中文版PV-Tech每日光伏新闻网站现已面市 PV-Tech.org 推出全新中文网站！ 敬请关注：http://cn.pv-tech.org 引导读者追寻光伏高新技术的源头！ 随时了解更多国际要闻？快来订阅PV-Tech中文电子周刊吧~ http://www.pv-tech.org/newsletter-cn/

新品发布

Photovoltaics International与PV-Tech.org是光伏行 业中稳居首位的国际信息机构，致力于为您提供光伏 领域中最权威的最新资讯与产品技术信息，是全球相 关企业决策者们的首选参考资料，其有印刷版和电子 版两种。

Sensors Unlimited高分辨率红外短波摄像机 | Sputnik研发MaxWeb xp数据记录器 | Schreiner ProTech压力补偿密封系统

Photovoltaics International 2010年第2期(中文创刊号)

光伏产业规模化进程

美国西图公司半导体产业规模化的经验借鉴

比利时研制高效率电池

应用微电子领域的经验

PV-Tech博客 First Solar 的确名副其实

本

公 司 W 诚 10 邀 馆 您莅 W 临 82 上 12 海 展 S 台 NE

C

我们的编辑顾问委员会成员均由国际太阳能领先企业 的决策者组成，如中国尚德公司的董事长施正荣就是 其中之一。这些业界精英给我们做后盾，确保了我们 的信息始终具有行业的前瞻性、针对性和最高水准。

www.pv-tech.org PVI_Lite_Chinese_Cover_02.indd 1

Photovoltaics International

lite

JobsinPV.com Recruitment

PV-tech.org Daily News

SolarLeaders Te l e v i s i o n

16/04/2010 14:16:51

PV DIRECTORY

联系我们: Photovoltaics International & PV-Tech.org | Trans-World House | 100 City Road | EC1Y 2BP | London | United Kingdom 电话: +44 207 871 0123 | 传真: +44 207 871 0101 | 电邮: info@pv-tech.org PV chinese.indd 1

23/09/2010 14:04:38

大略省的上网电价补贴政策实际包含 国际贸易法所禁止的补贴方案。

艾思玛太阳能再次调高全球 太阳能装机容量预期：预计 2010年可达17GW

Solarwatt AG公司首席运营官Ulrich Link博士为萨克森州州长Stanislav Tillich展示公司新生产线

新生产线。据公司表示，该生产线的 生产能力翻倍至400兆瓦，即每28秒便 可制造出一个太阳能组件。同时，公 司还建成一座占地13,000平米的物流 中心，并为其装配了260kWp的太阳能 发电设备。整个项目的总投资额高达 3500万欧元。 “通过建立世界最先进的 组件生产设备，并于今天开始生 产，Solarwatt有足够的信心和能力 迎接未来几年所可能出现的各种各种 挑战。”首席执行官弗兰克•施奈德 (Frank Schneider)表示。 根据西方制造商的标准，决定 组件制造商竞争力的关键因素是高度 自动化的工厂。据Solarwatt公司表 示，新生产线使用29个库卡系统(Kuka Systems)旗下的工业机器人，包括 Robo Frame，Robo Trimm以及Robo Load解决方案。 公司还表示，三班制的工作模式 将总计新增140个工作岗位。 市场动态

日本就安省上网电价补贴要 求本地产品配额一事向WTO 提出抗议 加拿大安大略省将使用一定比例 本地生产的产品作为享受入网电价补 贴的先决条件。这一做法招致了日本 的不满。根据路透社的消息，日本企 业已经向世界贸易组织(WTO)提起正式 申诉。双方将就这一问题展开会晤， 但最终可能还是需要进入WTO争端解决 程序来裁决安省的上网电价补贴要求 是否合法。 由于安省市场具有极强的吸引 力，供应链上多家光伏产品制造上已 宣布在安省投资建厂以符合其政策要 求。 欧盟法律禁止类似地方保护条 款，目前日本方面提出的观点认为安 8

w w w. pv -te c h .o rg

处于全球太阳能逆变器的市场领 军地位的艾思玛太阳能技术股份公司 (SMA Solar Technology)表示，由于 全球光伏市场的表现较之前预期有了 大幅飞跃，公司本年内第二次调高了 收入预期。SMA太阳能预计2010年全球 光伏市场新增装机容量将达到17GW， 比逆变器市场高出40%，这将给公司 带来17～19亿欧元的收入。而此前公 司对这两项数据的预测分别是14GW和 15～18亿欧元。 SMA太阳能表示德国还将保持其最 大单一国家光伏市场的地位，今年新 增装机容量将达到8GW。南欧和美国市 场也在加速增长。 不过，公司表示对2011年的增长 很难做出预测。 一方面，2011年全球光伏市场预 计增长20%；另一方面，各国对上网电 价补贴政策的调整和其对市场潜在的 负面影响可能导致光伏市场在明年下 滑10%。 美国和意大利被认为是决定明年 增长幅度的关键国家，两国都是明年 增长的主力。

加州拟通过新刺激计划大力 促进中型太阳能项目 加州公用事业委员会(California Public Utilities Commission，简称 CPUC)近日公布一项新提议，决定实行 一项旨在促进中型可再生能源项目发 展的新刺激计划。此项新上网电价补 贴政策要求由各投资者控股的加州公 共事业单位从规模在1至20MW之间的可 再生能源设备处购买电力。 负责与CPUC共同实行该新政策的 Vote Solar公司执行董事亚当·布朗 宁(Adam Browning)表示：“加州政府 的政策对于开发大型公共事业规模太 阳能电站和发展小型家用商用系统来 说是强有力的后盾，但是，这两种规 模之间有着不小的落差。CPUC此次的 提议旨在填补此空白，为太阳能市场 的发展、就业岗位的增加提供又一良 机，并在短时间内为加州新增了大量 的清洁能源。” “太阳能政策应可通过提高工 作透明度、提供公平的竞争环境和可 靠的市场机遇，为基金会提供长期的 市场发展。”来自新政策另一支持者 IREC的法律代理公司Keyes & Fox的凯 文·福克斯(Kevin Fox)表示，“此计 划通过新型定价机制达成了许多大型

艾思玛太阳能技术股份公司的Sunny Boy 逆变器

政策目标，同时还保证了加州纳税人 的利益，并克服了美国市场所存在阻 碍上网电价政策发展的法制方面的困 难。” CPUC在提议中表示，将率先在符 合条件的中型可再生能源设备中选出 1GW作为该政策的实验性实施对象。这 就要求加州最大的三家由投资者持股 的使用单位每年针对可再生能源开发 商进行两次招标活动。政府还要求各 事业单位必须按照投标价格由低至高 地分别与各开发商签署合同，直至当 轮招标项目量全部完成为止。 此计划将使用标准化条款以降低 交易费用，并为高效融资提供所需相 应的合同透明度。所涉及项目的可行 性将由安全开发机制以及相对较短的 项目开发时间表保障。CPUC将对此计 划作出最终裁定并最快将于三十天内 正式实施此计划。 电池技术

晶澳太阳能供签署2011年 500兆瓦供应协议 晶澳太阳能控股公司近日与几家 未透露姓名的客户签署了多个供应协 议，这将保证该公司在2011年供应超 过500兆瓦的单晶及多晶太阳能电池。 协议称供货日期起始于2011年1月，直 至2011年12月，并已为次年应交付的 太阳能电池采取了预付款的方式。另 外，据晶澳太阳能透露，其2010年的 年终生产能力将达到1.8GW，第二季度 终期实际生产能力为1.4GW。 “我们从全球多样化的顾客群中 看到了他们对于我们先进技术、高质 量产品的强劲需求。”晶澳太阳能的 首席执行官方朋博士表示，“我们的 顾客视晶澳为一个令人信赖的供应伙 伴，并相信晶澳能满足他们2011年的 生产需求。我们很高兴能为明年的签 署这些新的供应合同，这也进一步明 朗了我们2011年的前景，并充分表明 了我们依然保持着工业供应高质量太 阳能产品的领导地位。”

企业： 中电光伏有限公司

100 MW

晶澳太阳能有限公司

150 MW

无锡尚德太阳能电力有限公司

150 MW

阿特斯阳光电力公司

50 MW

昱晶能源科技股份有限公司

50 MW

江阴浚鑫科技有限公司

50 MW

正泰集团股份有限公司

50 MW

新日光能源科技股份有限公司

50 MW

总额：

650 MW

卡姆丹克2011年硅片供应协议

卡姆丹克太阳能将在2011年 交付超过600MW太阳能级单 晶硅片 卡姆丹克太阳能系统集团有限公 司及其子公司近日与主要合作伙伴签 署了价格待定的晶片供应框架协议。 卡姆丹克将向包括昱晶能源科技股份 有限公司、江阴浚鑫科技有限公司、 正泰集团股份有限公司及新日光能源 科技股份有限公司在内的各主要合作 伙伴提供总计200MW的太阳能级单晶硅 片。此外该公司已与中电光伏、晶澳 太阳能、尚德太阳能和阿特斯阳光电 力签署了供货协议。 根据协议，卡姆丹克将在2011年 1月到2011年12月间分别向昱晶、浚 鑫、正泰及新日光提供大约50MW单晶 硅片。 “我们很高兴看到在主要客户强 劲需求的刺激下，2011年的订单已经 突破2010年底预期产能的600MW。目前 我们还无法满足客户的需求，而客户 还在不断追加2011年的出货订单。” 卡姆丹克的董事长兼首席执行官张屹 表示。 “在2011年底前我们的产能将会 提升至1000MW，我们有信心填补产能 上的不足。目前，我们的主要任务是 加快400MW的产能扩张计划。”张屹总 结道。

江西赛维LDK在新余市完成 首条太阳能电池生产线 江西赛维LDK太阳能高科技有限公 司已于近日成功完成了其位于新余市 的首条太阳能电池生产线的安装调试 并进行试运行。这条太阳能电池生产 线年产量目前为60MW；预计在2010年 第三季度末产量将达到120MW。 “我们很高兴能宣布公司已正式 开始了太阳能电池产品的生产，”赛 维LDK的董事长兼首席执行官彭小峰表 示，“我们的团队在最短的时间内成 功地完成了新生产线的试生产，这是 我们在奉行公司一体化增长战略的道

路上取得的又一个重要里程碑。我们 内部电池生产的目标是要满足我们组 件生产50%的需求，以此来降低组件 成本，同时保证电池的稳定供应。我 们在新余市新建的太阳能电池生产设 施进一步贯彻了这一战略。” 原料供应

英利与OCI Chemical签署为 期5年的多晶硅供应协议 英利绿色能源控股有限公司与韩 国领先的化工生产商OCI Chemical签 署了一项为期5年的多晶硅供应协议。 根据协议，OCI同意在2011年到2015年 间向英利提供总价格为4.42亿美元(约 合30亿人民币)的多晶硅。 2010年第二季度的英利在销售 收入、毛利润和净收入等方面表现强 劲，各项指标全面上涨。这家中国纵 向一体化太阳能产品制造商表示，其 在6月30日前的一段时间内，出货量经 历了一波较大幅度的增长，并预计全 年的出货将达950MW到1GW。

签订了一份独家的多年协议，将在后 者的生产设施内提供现场回收处理服 务。预期至2011年底，Nexolon 的太 阳能硅片产量将超过15亿瓦特。届 时，Nexolon将成为全球最顶级的太阳 能硅片制造商之一；目前，其已经是 韩国最大的供应商。 “Nexolon对与CRS的合作前景充 满信心。该公司在帮助业务合作伙伴 增加回收率、降低成本方面拥有良好 的声誉，”Nexolon 战略计划部总监 Sung Joon Kim 表示。“CRS的计划将 使用多样化的切削液提供始终如一的 优质砂浆，同时还尽可能地减少了沾 污问题，这正是我们所需要的。” “CRS的独特现场模式和他们同其 他业务合作伙伴之间的紧密关系使我 们坚定了与其签订长期合作协议的信 心。我们期待在这一领域取得持久的 成功。”Kim补充道。 该工厂位于韩国益山，将利用 CRS的统包解决方案来回收处理光伏产 业中生产太阳能硅片所用的砂浆。合 同有效期为五年，分四个阶段扩建。 生产高峰时期，预计砂浆的使用量将 达到每年36,000公吨。

德山化工马来西亚Samalaju 工业园区多晶硅新厂建造计 划正式启动

OCI Chemical制造厂

Silicon Genesis将PolyMax系 统用于太阳能硅片生产 Silicon Genesis(SiGen)公司已 在使用PolyMax系统进行太阳能硅片生 产方面取得了一定进展，并认为该工 艺可取代传统线锯工艺。SiGen已经生 产了厚度为85微米，面积为156平方毫 米的无切损单晶硅晶片。切损是切锯 过程中用于转化为锯屑的材料。 SiGen的首席执行官弗朗索瓦•亨 利(Francois Henley)先生表示：“ 我们相信无切损硅片的使用将促使光 伏产业在无补贴的条件下达到电网平 价。而公司旗下制造系统的量产则是 实现这一目标的关键。” SiGen将在于西班牙瓦伦西亚召开 的第二十五届欧洲光伏太阳能会展上 推出其PolyMax系列产品。

尽管公司于2008年11月份就 开始了厂址的筛选工作，德山化工 (Tokuyama)直至今日才正式在马来西 亚Samalaju重工业园区建造6000吨多 晶硅工厂。据公司表示，该工厂将于 2011年初正式动工，并预计于2013年 春季投产运营。此前公司曾宣布该厂 的建造费用将达650亿日元(约51.9亿 人民币)，而设备花销将达150亿日元( 约12亿人民币)。 该厂将致力于像太阳能产业供 应其产品。此前，德山化工曾计划在 2012年实现该厂投产运营，而太阳能 及半导体产品的额定产量也仅为3000 吨。

CRS与Nexolon签订长期回收 处理服务合同 CRS Reprocessing Services (CRS)近期与 Nexolon Company, Ltd.

CRS 的浆料回收系统

Photov ol tai c s I nter nati on al

9

杜邦專家觀點 太陽能(PV)背板材料效能分析比較

太陽能背板材料必須長效且可靠，才能確保太陽能模組能維持一 貫的電力輸出及物理完整性長達25年以上。 畢竟，既然模組製造商保證其模組可以運作25年以上，那麼他們 就得對所選用的材料相當有信心，確定這些材料不會隨著時間而 使模組效能喪失或下降。 作為全球太陽能產業材料的頂尖供應商，杜邦知道選用材料的重 要性。因此，我們目前正在進行大規模的測試，深入研究材料效 能會如何影響模組效能，以及其最終能帶給模組的益處。 Michael DeBergalis 博士為杜邦太陽能解決方案的科技總監，是長 期進行太陽能背板測試的關鍵技術專家。他已在杜邦任職32年， 並持續進行Tedlar®聚氟乙烯(PVF)薄膜的創新發展超過20年，同 時在過去9年中一直致力於研發適合太陽能產業的高效能背板，並 且持續在全球各地的研討會中發表刊物及演講來分享他的觀點。

Dr. Michael DeBergalis 杜邦太陽能解決方案之科技總監

最新的材料測試結果，已於2010年9月在西班牙瓦倫西亞所舉行的 「歐洲太陽能源研討會」中發表。以下為重點摘要，可讓我們對 這項研究有初步的了解。 目前杜邦測試過哪些類型的背板結構？ 杜邦已對市面上使用最廣泛的商用太陽能背板進行測試，包括一 些Tedlar®背板結構及數個熱門的材料結構。大多數的背板為多層 複合物，可多方面強化太陽能模組的效能。它們能提供保護、隔 離環境傷害、提供電力絕緣及兼具美觀功能。背板要能讓模組的 整個生命週期擁有上述的特質，而目前的發展可讓以上特質維持 25年以上。由於Tedlar® PVF薄膜可呈現高效能背板所需的特質， 讓「TPT™」(Tedlar®/ Polyester/Tedlar®)成為目前最受歡迎的背 板之一。在這些背板結構中，Tedlar®薄膜扮演著雙重角色。其外 層可隔離環境傷害，保護模組，而內層能長期有效地與模組封裝 材穩固黏著，同時提供保護作用。

10

w w w. pv -te c h .o rg

為何選擇搭配Tedlar®薄膜的背板？

杜邦的發現為何？

主要考量原因是效能與設計。由於目前常用的「玻璃/背

詳細的發現成果，將會於2010年9月在西班牙瓦倫西亞所舉

板」模組是於1970年代設計的，許多材料都被考慮用來作 成聚合物的背板材。相較於其他可用的材料，Tedlar ® 薄

行的「歐洲太陽能源研討會」中發表。杜邦的3個主要發現 為：

膜它同時擁有耐久、黏著、色美及價廉等特性，因此選用 Tedlar®薄膜乃為最佳選擇。這些特性至今依然相當重要。

第一，在溼熱的環境測試背板對封裝層的黏著度，1000個

當你想到以上特性，以及它的25年超長效能時，Tedlar®自

小時後，我們發現部份以PET與氟素高分子為主結構的背

然而然成為最佳選項，且相較於其他材料，它一直是這個產

板出現黏著力或機械力喪失的情形。而TPT™結構則完全相

業的標準用料。

反，在曝曬後它的黏著力甚至更加提升。

杜邦選擇檢驗的材料特性，以及檢驗的理由為？

第二，在溼熱的環境測試背板的形變程度，我們發現PVDF

我們在相同結構及環境的同步比較基礎上，測試了各種背板

背板在500個小時後發生脆化，並且在1000個小時後更加惡

特性、執行一連串的曝曬測試以及視覺檢測多種背板材料。

化。而TPT™結構在曝曬後的效能依然相當優異穩定。

我們測試的物理特性包含如下： · 對乙烯醋酸(EVA)的剝離強度－決定背板對封裝層的黏 著度。

第三，在進行顏色變化測試時，我們發現全聚酯材背板與

· 張力特性－決定材料的物理完整性、機械強度及抗裂 性。

現重大改變。相反地，TPT™背板僅出現些微的顏色改變。

PFAVE背板在氙氣曝曬1190小時及2380小時後，顏色皆出

· 絕緣破壞電壓－以安全性及效能來看，該特性可說是相 當重要。 · 熱膨脹係數－決定材料曝曬在溫差大的環境下所受的影 響。 · 顏色/顏色穩定性－曝曬造成材料顏色改變，可能會導致 材料分解及效能降低。 · 透濕防水性－決定模組在潮濕環境中吸收水分的功能。 · 剪切性－決定背板是否會在處理時毀損。 · 反射係數－可以透過面板間光線的反射來提升模組效 能。

本次研究的主要發現為？

· 尺寸穩定性－確保背板材料能在製造過程中與模組相 容。

時將牽涉到的各種特性納入考量。在加速條件下測試背板的

我們了解在設計太陽能模組及選擇適當的背板材料時，須同 特性是否穩定，才能確保背板能夠長久耐用。我們所面臨的 挑戰主要為了解材料效能及模組效能間的關係。這類的研究

爲了評估以上特性，杜邦執行了哪些測試？

能協助相關人員選擇合適的背板材料，以確保模組的長期效

我們採用相當具代表性的國際電工學會(IEC)、美國再保險協

能，亦能協助我們的新產品開發團隊，讓他們在這塊變化莫

會(UL)及美國材料試驗協會(ASTM)的測試條件，來執行所有

測的產業，持續提供最有效的解決方案。

測試；同時，為了取得更長久的曝曬度，我們還執行高於標 準要求的加速曝曬測試。

下次預備進行的測試為何？ 緊接著材料測試後，我們預備進行模組及加速老化測試。同 時我們也將進行戶外實績的測試。 很多搭配TPT™背板的PV模組已經順利運作長達25年 了。Tedlar®是唯一能提供模組長達25年效能的背板材料， 也無疑地能繼續在未來的日子裡，提供耐久可靠的電力輸出

tedlar.com/backsheet

及模組效能。

Copyright © 2010 DuPont. The DuPont Oval Logo, DuPont™, The miracles of science™, Tedlar ® and TPT™ are trademarks or registered trademarks of E.I. du Pont de Nemours and Company or its affiliates. All rights reserved. 05/10

Photov ol tai c s I nter nati on al

11

新品发布 伊斯拉视像公司推出高新多面板光学检测系统 产品应用：该产品可针对压层板进 行检测，检测内容包括传入玻璃、 边缘破损、薄膜检测(如透明导电氧 化物等)以及划痕检测。 相关信息：伊斯拉公司在整个产品 开发过程中启用了自主设计并制造 的标准化组件及其零部件。此举保 证了公司可在最大程度上拥有客户 群的灵活性。据称，该系统还拥有 较高程度的兼容性以及操作安全 性，并能列出重复性的错误、可实 时显示产能，同时其坚固程度和可 靠程度也十分值得嘉许。 上市时间：2010年3月

伊斯拉视像公司(ISRA Vision)推出了 一款名为Powerscan的高新多面板光学 检测系统，用于光伏业内玻璃、表面、 压层版等的自动检测。在其标准化的组 件和应用软件的基础上，伊斯拉公司开 发出一套系统设置，用于检测完成压层 工艺后的薄膜太阳能组件。Powerscan 在压层工艺检测上的应用主要是针对压 层过程中所产生的气泡、裂痕、PVB过 量、边缘破损以及玻璃自身的划痕等问 题。凭借相机、光照及边框等模块化产 品构架，伊斯拉公司基本上能够胜任所 有的检测工作。此外，该产品还具有人 性化操作界面、检测批次上的灵活性， 以及在薄膜太阳能业内检测任务完成水 平较高等特征。

TwentyNinety公司MLC技术可大幅改善光伏阵列性能 产品应用：可覆盖多达8组Active Array组件 相关信息：Active Array的核心技 术在于Active Tag。Active Tag是 一款集成在每个光伏组件接线盒上 的低成本小型设备，用于监控组件 的性能和状况，并通过无线网络将 数据传送至Active Combiner进行组 件及行列的数据分析，同时，该设 备还可对危险情况作出反应，或对 轻微性能及安全问题想操作人员进 行示警。Active Combiner是一款 多通道设备，可对多达8列Active Array组件进行管理，并有两个完整 的智能反向输出电流保护装置。 上市时间：已上市。

TwentyNinety公司推出的Active Array 系列组件监控技术(MLC)可实现低成本 无线运行，以改善光伏组件阵列的性能 并提高其安全系数。此系统可对光伏组 件性能进行跟踪记录、及早发现问题、 优化系统结构并保证数据监控，还可发 现并隔离运转不良或被遮挡的组件，以 提高整体阵列的性能。此外，系统对组 件的智能监控可通过无线运行方式覆盖 整个光伏阵列，从而实现对潜在问题的 早期诊断，并在无需接触危险操作环境 的情况下远程绕过个别运行不良的组 件。在紧急情况下，系统操作人员或消 防人员可远程对个别组件、组件列甚至 是整个阵列进行隔离。此外系统还可自 动隔离由接线盒内电力问题而引起的超 温组件，即时降低火灾风险。

坎贝尔公司为大型太阳能设备提供双轴跟踪系统 产品应用：此系统可用于大型太阳 能发电设备，并可根据不同设备进 行客制生产。 相关信息：此系统的防翻转、热浸 镀锌钢架结构可抵挡所有等级的风 力，其雪荷载也已达到DIN 1055水 平。公司还可根据客户要求为其配 置超强抗风防护装置。此系统十分 耐用并已通过业内标准测试。系统 所配备的中央控制系统可兼容组件 的外延设备，如浪涌电压保护器， 并可经由充电电池提供应急电力。 该电池还拥有一套网络评估软件， 用以监测器充电状况。 上市时间：2010年4月 坎贝尔公司( K e m p e r )4月份推出

12

w w w. pv -te c h .o rg

的KemTrack120跟踪系统，可覆盖高达 120平方米的组件面积。由于设备占地 面积的限制，以及对特定地面支架光伏 项目输出量最大化的要求，大型高效的 跟踪系统已成为众多应用系统运行时不 可或缺的设备。因此，抗风力强、雪荷 载高的坚固持久的架构也成为了跟踪系 统的必备条件。目前该型号系统的单位 系统产量为17KWP，而KemTrack60系统 的单位系统产量仅为8.6KWP。此外，该 系统还可通过标准规格的集装箱在短时 间内在全球范围进行运输，不仅节省了 空间还节省了与运输相关的各种费用。 坎贝尔公司表示，此新设备中的两套天 文调控系统可将运行效率在现有设备的 基础上提高40%。系统中加装的软件算 法可通过阳光辐射角度防止设备出现相 互遮挡的情况。

蓝星有机硅旗下封装系列产品可在不同温度下保持其弹性 产品应用：本品可用于光伏接线盒及 包括热太阳能面板等组件产品的密封 与黏接，并同时具有固化速度快、机 械性能高、对于金属及塑料的自粘附 性和非腐蚀性等优点。 相关信息：相关信息：蓝星有机硅 公司已根据IEC61215标准对CAF粘合 剂进行了老化测试以检测产品在PVF 和合成PVDF/PET合成材料上的粘合性 能。公司对产品进行了湿热测试，即 将产品置于相对湿度为85%的85°C环 境中长达1000小时。另一项热循环测 试是将产品置于-40°C至+85°C的温 度变化范围中。此外，公司还对其产 品进行了搭接剪切测试，以测量产品 的搭接剪切强度与粘度极限。 上市时间：2010年3月

蓝星有机硅公司(Bluestar Silicones) 近日推出的“CAF 530”系列产品可通 过提高组件在运行环境下的抗腐蚀性对 组件的高性能使用期进行优化。据称该 公司的技术可保证组件在特定的环境 内正常运行至少25年。蓝星有机硅的 CAF530粘合剂拥有高度弹性，能够抵 御并吸收组装材料间的示差膨胀。此系 列粘合剂的成分包括单组分及双组分硫 化硅橡胶(即RTV 1 & 2)，用来对组件 框架及接线盒进行粘结和密封，同时， 双组分产品还可用来对接线盒和光伏电 池中的零部件进行封装。CAF粘合剂是 一种室温硫化硅橡胶(RTV-1)，可在室 温条件下，在产品与大气湿度的瞬间固 化。其固化速率还取决于环境温度与湿 度的水平。

AEG公司Thyrobox M能源解决方案有效降低多晶硅每千克生产能耗 产品应用：产品可用于多晶硅制造 中的沉积工艺中。 相关信息：AEG公司所设计的系统解 决方案能够很好地与电网将连接， 并可在恶劣的环境中为关键任务应 用设备提供能源解决方案，包括发 电站、海上石油钻井平台、化工厂 和公共事业规模可再生能源电站 等。 上市时间：已上市。 AEG公司推出的Thyrobox M系列产品， 据称是目前最集中化的能源解决方案， 可使得多晶硅制造中的沉积工艺更加稳 定可靠。除了拥有其先进的流程监控能 力、高达0.95的功率因数以及超过99% 的能源效率，Thyrobox M还能减少整个

能源供应过程中安装、调试及维护的成 本，并且，由于该产品占地面积要比上 一代产品及其他竞争公司所推出的产品 要小得多，因此设备的安装成本也要低 得多。该产品的占地面积比上一代产品 少25%，同时也比其他竞争企业的产品 节省了近70%的占地面积。这种集中化 的设计可有效减少还能减少整个能源供 应过程中安装、调试、维护及设备本身 安装的成本。其预警系统和故障排除功 能可将由于硅棒断裂融化所引起的生产 中断风险最小化。而设备中集成的中压 点火器也提高了制造流程的稳定性。此 外，Thyrobox M产品还可通过监控系统 对能源消耗及硅棒重量进行实时数据收 集与分析，以增加能源使用效率。

欧瑞康莱宝真空设备推出DRYVAC真空泵,涵盖多种处理设备 产品应用：可大规模应用于晶硅及 薄膜产品的制造流程中。 相关信息： D R Y V A C 真 空 泵 系 统 及“-i”系列产品包含外壳、脚 轮、变频器，可编程遥控器和触摸 屏等标准化配置，并可通过菜单导 航、软件和现场总线等直观技术实 现控制监测的可视化。 上市时间：已上市。 欧瑞康莱宝真空设备公司推出了DRYVAC 系列新款干式真空泵。集合了多种创新 性设计，包括涂层过程和各种工业制造 流程中的特殊要求。在降低成本的进 程中，低能耗占据了越来越重要的地 位。然而，随着制作工艺的不断发展 和变化，拥有高系统运行时间的小巧

灵活的集成产品也同时成为业内研发 的目标。DRYVAC系列中的Sprinter和 Enduro产品已实现了在多种压力条件 下抽气速度最优化，范围由0.02毫帕 至一个大气压强。这就使得这些产品 能够实现快速抽空，并十分适用于真 空传送设备。Enduro产品据称可完全 满足工业流程中对相关设备的所有基 本要求。DRYVAC Champion系列的产品 是为保障苛刻运行条件下的设备稳定性 而特别设计的。这些真空泵已根据光伏 及平面面板制造过程中所产生的气体而 进行优化改良。DRYVAC系列产品包括以 下型号：DRYVAC 650 S、DRYVAC 650 S-I、DRYVAC 5000 RS-I，以及与RUVAC 系列的鼓风机交叉组装的多系统组合。

Photov ol tai c s I nter nati on al

13

多晶硅硅片制造流程上 的降本增产策略

马克·奥斯本(Mark Osborne), Photovoltaics International 高级新闻编辑

此技术文章节选自第八期印刷版《Photovoltaics International》季刊

摘要 目前光伏市场中有约60%的多晶硅太阳能电池是以多晶硅硅片作为基底的，因此，多晶硅硅片可谓是光伏产业的 核心主力。多晶硅硅片由于有较高的转换效率，在价格上占足了优势，并且这一优势还将随着电池技术的发展而继续 持续下去。但事实上，从2008年下半年开始并持续了2009年大半年的光伏市场需求的下降，确实对硅片价格造成了冲 击。面对资本成本相对较高、价格压力不断增加以及对质量和控制的要求提高，硅片制造商正努力实行严格且可持续 的生产成本削减战略，以满足客户的各种需求。本文将着重探讨可用于实现更严格的质量标准的策略，以及如何降低 下游厂商的生产成本、提高电池转换率等问题。

前言 在过去的2009年里，多晶硅、 硅片，以及组件的价格急遽下跌。据 iSuppi公司最近的一项调查显示，晶 硅组件的价格下跌了37.8%，同时，太 阳能硅片的价格下降了50％，而多晶 硅价格的降幅则达到了80％。 这家市场调研公司预测，2010年 相关产品的价格仍将进一步下跌，但 降幅将远小于09年。在谈及2010年的 价格走势时，iSuppli公司认为晶硅 组件的价格将下降20％，太阳能硅片 和多晶硅将分别下降18.2%和56.3%。 （见图1）许多业内观察家都认为， 随着业内主要制造商及各新兴制造商 对各自多晶硅产品的进一步扩产，涉 及业内整个供应链的降价情况已成定 局。 当然，对于一个追求难以实现 的电网平价，甚至更低价格的产业来 说，原材料降价是实现目标的重要前 提条件。而且只有实现了电网平价， 才能赋予太阳能比其他形式的可再生 能源更大的优势，从而使从业者从中 获益。

“价格的下跌必将改变太阳能产 业的现状，”iSuppli公司光伏系统 高级总监兼首席分析师亨宁·维希特 (Henning Wicht)表示，“光伏产品价 格的大幅下降意味着价格结构正经历 着一场不可逆转的调整，而这种情况 必然会促使整个产业向更具竞争性的 市场形态转型。” iSuppli公司认为，为了应对产品 的大幅降价，各光伏产品制造商应继 续将精力放在节省成本方面，以弥补 其2009年所遭受的利润损失。维希特 先生还指出，当成本下调的幅度最终 与价格下降的程度相一致时，整个产 业的总体利润水平将可得到改善。 这对于多晶硅和硅片制造商来说 是十分重要的。尽管他们曾经历价格 的剧烈下跌，但是比起那些纯粹组件 制造商这样的下游企业来说，也具有 相当高的资本成本。 当然，硅片制造商也同样要面对 各种挑战。随着市场对硅片的强劲需 求创造出一个相对严峻的供应形势， 若想应对客户的需求，进一步的资本 扩张势在必行。通常情况下，在市场

需求强劲时期，设备的交货期将延至 九个月以上，因为像铸锭炉和线锯等 关键设备的生产周期都相对漫长。大 多数这些设备规模庞大、工艺复杂， 因此很难在较短的交货时间内保证制 造质量。

多晶硅的重要性 日前，多晶硅的短缺造成了供应 链上的断流，而这种危机很有可能向 硅片产业转移。这就使得整个产业在 计划增产并采取实际行动的同时，需 大力发展各自的生产力。 正如图2所示，iSuppli公司对硅 片制造商的增产计划做了调查，实际 结果表明，硅片的产量只有小幅上涨( 左柱)。中柱表示那些本应实现却未付 诸实践的增产计划。只有在2012年及 之后的年份里，那些增产计划才有可 能大规模实现。 可以说，这样一份报告仅仅是再 次为硅片制造商们带来压力、促使其 优化现有的产品吞吐量、出货量和总 体产量而已。新技术的引进同样可以 提高产量，同时还可降低生产成本。

小型系统（$/W）

晶硅电池（$/W） 硅片（$/W） 硅（$/kg，合同） 硅（$/kg，现货）

图1：2008-2012年销售均价预测（US$/MW）

14

w w w. pv -te c h .o rg

图片来源：iSuppli公司

平均价格（$/kg）

美元

晶硅组件（$/W）

硅片产能计划 硅片增产计划

图片来源：iSuppli公司

硅片总产能（iSupply预测）

图2：全球太阳能硅片产量计划（MW）

那些有能力成功应对这些挑战的企业 不仅能凭此降低成本，从中获利，还 能争取到那些需要以低成本来维持竞 争力的下游客户。而那些下游企业也 能从总体价格不断下降的大环境中获 益匪浅。 毫无疑问，多晶硅和太阳能硅 片的关系密不可分。硅片价格之所 以下跌，在很大程度上是由自08年中 期起出现的原生多晶硅价格下跌所导 致的。许多市场调研报告均指出，持 续的产能扩张将导致价格的进一步下 跌。Bernreuter信息研究公司近日所 公布的一份报告指出，至2012年止， 全球多晶硅产量将达25万吨，其中仅 中国就可生产8万吨，占到全球总产量 的近三分之一。

Bernreuter信息研究公司创始 人约翰内斯·本罗伊特(Johannes Bernreuter)在接受本刊采访时表示， 今年多晶硅的供需动态平衡将使得产 品的现货价格浮动在每千克45至50美 元之间。 本罗伊特先生还表示：“就我个 人认为，在2011年里，中国制造商的 大规模产量将很可能将生产成本降至 每千克35美元以下。同时也会使现货 价格降至相应水平。这种情况最快将 会在年底，甚至更早出现。”因此， 硅片价格将会进一步下跌，但正如我 们在2009年里所经历的那样，产品降 价的负担将远大于生产成本，从而导 致了边际利润额的大幅缩水。但从另 一个角度来看，那些硅片制造商实际

关于铸锭炉 ——尺寸是有效之道吗？

公吨

低供应 价格下降 高需求

供应 需求

年份 图3：Bernreuter Research：多晶硅供需价格分布

16

w w w. pv -te c h .o rg

上能够从多晶硅的多渠道供应中获 益。曾任赛维太阳能公司(LDK Solar) 生产副总监的产业顾问尼克·萨诺 (Nick Sarno)认为，目前市场上可获 得的原生多晶硅的质量已越来越好， 并且硅锭的制造过程对硅废料的依赖 性也越来越低，因此，业内材料质量 在总体上得到了提高，从而使硅片制 造商从中获利。 “这就可以在硅片制造流程的 各个阶段进行成本下调。”萨诺先生 还表示，“缩短在硅废料采购、分类 及储存阶段的工时并不会对整个生产 流程造成很大影响，但优质原生材料 的供应可改善硅锭、硅片及硅块的质 量，减少废料的产生，降低消耗品的 使用量，从而全面提高产品质量。”

为了降低硅锭成本，增加硅锭 尺寸逐渐成为一种主流趋势。硅锭 尺寸越大，其所制成的硅块就越多， 从中就能够制造出更多的硅片，进而 可提高总体产量(见表一)。这就需要 大尺寸的硅锭生长熔炉。目前，大多 数硅锭的生产使用的都是定向凝固 系统(Directional Solidification System，简称DSS)铸锭炉，该系统可 使用450kg的铸锭炉对多晶硅硅锭的铸 造进行量产。500kg和800kg以上规模 的铸锭炉制造技术也已得到了长足的 发展，并且，在不远的将来，1000kg 以上的铸锭炉定将会在整个产业不屑 的努力下成为现实的。 目前市场中，铸锭炉尺寸发展的 最大制约是无法在技术上实现大尺寸 坩埚的制造。通常情况下，业内企业 应对此问题的方式是在大型铸锭炉内

产品型号

JZ-270

JZ-450

JZ-520

JZ-800

多晶硅初始消耗（kg）

270

450

520

800

680

840

840

996

250

270

315

345

156

156

156

156

200

220

265

295

锭的预计切割重量（kg）

181.4

311.8

375.7

602.2

锭的预计产量

67.2

69.3

72.3

75.3

凝固锭预计尺寸（毫米） 长、宽 厚度 裁剪块预计尺寸（毫米） 长、宽 厚度

表一：JYT大型熔炉效益分析

同时使用多个坩埚，以实现设备的最 优使用和产量的增加。 “从资本扩张的角度来看，这 一策略的好处很多。”萨尔诺先生表 示，“这一策略不仅可以适应大型多 坩埚制造实现后的投资形式发展，同 时由于坩埚尺寸的限制而无法扩大铸 锭炉可有效防止过度投资。” 坩埚制造商正试图同时解决多个 问题。除了要满足对大尺寸坩埚不断 上涨的需求，其中以450kg的坩埚最受 欢迎，还同时为未来制造更大尺寸的 坩埚进行设备更新。 石英坩埚制造业龙头赛瑞丹公司 (Ceradyne)已于2010年上半年对其赛 瑞丹(天津)工业陶瓷的工厂进行增产 计划，同时，旗下的赛瑞丹(天津)先 进材料公司的新厂区也已破土动工， 预计将于2011年初期实现运营，以满 足不断上涨的市场需求。 GT Solar公司是DSS铸锭炉领域 内的主要供应商，该公司在过去的几 年里对市场面对大型设备不断上涨的 需求有着充分的了解。在2009及2010 两个财政年度里，GT Solar公司共向 市场售出近10亿美元的资本设备，用 于为已有终端用户增建设备约8GW至 10GW。 有趣的是，GT Solar公司光伏

设备部门产品营销总监周先生(Henry Chou)在交谈中所提及的首要问题却并 不是扩大铸锭炉尺寸的问题。周总监 所谈及的第一件事是，随着产业市场 的日趋成熟，GT Solar公司所关注的 是其称为“价值尺度”的问题。通过 使用简化的变量来强调生产成本，公 司期望凭借提高制造工具的吞吐量来 降低成本，以获得较高的产量以及较 好的硅锭总体质量。这种方法可以在 从硅锭至硅片切割的整个流程中减少 浪费情况，同时降低成本。 “事实上，获得较高质量的硅片 并不意味着电池制造商能够获得较高 的电池转换率。”周总监还指出，“ 我们所寻求的是每小时所生产的千克 数。” 周总监还表示，衡量铸锭炉成本 下调的两大关键指标是铸锭炉的吞吐 量及产量。其中，吞吐量与铸锭炉在 单位处理时间里的熔料量有关。 “有趣的是，如果炉料的分量开 始大幅增加的话，所花费的处理时间 也将随之增加。根据处理时间长短的 不同，分离杂质的质量以及晶体的生 长可改变整体的最优平衡。因此，由 于晶体生长周期也相应延长，即使铸 造出大型的硅锭，也并不意味着能获 得更高的吞吐量。”

图4：赛瑞丹（天津）工业陶瓷有限公司的钳锅

周总监还认为，如果硅锭尺寸超 出了600kg，就很可能会导致对其他 设备进行调整，从而在事实上增加了 生产成本。使用超过450kg的熔料量 将成为不可避免的趋势这一结论看起 来并不成熟。该公司近日凭借旗下的 DSS450铸锭炉及其辅助设备和相关服 务，获高达2亿美元的资产估算额。 众多硅片制造商与其签署了订单，其 中包括天威新能源、Phoenix光伏科 技公司、英利绿色能源公司、晶澳太 阳能(JA Solar)和中美晶硅产品公司 (Sino-American Silicon Products， 简称SAS)。 在2010年5月初的时候，GT Solar 推出了一款新产品DSS450HP高性能硅 锭生长炉，该铸锭炉拥有经过热优化 的第二代热区技术。据称该技术可保 证在提供业内领先晶硅质量的同时提 高设备吞吐量。DSS450和DSS240两个 机型的用户可通过加装现场升级组件 将设备改装成新型DSS450HP设备。这 就意味着，硅片制造商将更倾向于在 短期内选用已成熟的技术来扩大其硅 锭/铸锭炉的尺寸。

浆料解决方案 浆料是硅片制造的核心。浆料常 被与线锯配合使用，用来将硅锭切割 成硅片。通常情况下浆料是一种由碳 化硅和乙二醇混合而成的磨料。与其 说使用浆料来切割硅锭，不如说是腐 蚀硅锭来得准确。在大多数情况下， 当谈及整个硅片制造流程中哪一步骤 是节约成本最关键的环节时，浆料从 来都是当之无愧的第一名。 当浆料沿着线锯的准绳流至硅锭 上时，其中的腐蚀性碳化硅(即沙砾即 可将硅锭切割成硅片)。浆料应该经常 更换并进行处理，以保证所产硅片的 数量和质量，并同时将单位成本维持 在可接受的范围内。 尽管浆料对于硅片生产流程来说 十分重要，其在生产成品中所占的比 例也不容忽视。浆料的成本通常仅次 于对晶硅本身的成本花销。尽管新型 浆料的价格相对较低，仅为每千克3.5 美元，但总体成本升高得却很快。假 设一条生产线上使用了10台线锯，其 运营成本可高达每年1600万美元，而 这其中还不包括废料排污费和工人劳 务费。 对于大多数制造商来说，浆料的 再处理高居各运营成本中的前五位。 对此项业务进行优化可在硅片生产成 本和总体质量等方面获得显著且长久 的成就。 据业内领先的浆料回收公司CRS Reprocessing Services 表示，正准 备开展浆料再处理的项目或已开始使 用新工艺的硅片制造商能够借由边际 效益的大幅改善而获益匪浅。以10台 Photov ol tai c s I nter nati on al

17

图5：GT Solar 生产的DSS450HP型锭生长炉

线锯同时运行、硅片月产量385吨为 例，若开展浆料再处理业务，每年可 在净回收成本、公用事业和基础设施 投资上节省800至1000万美元。 “在CRS公司，我们能够回收80% 至90%的沙砾及硅片承载器。”CRS公 司创始人兼总裁比尔·劳伦斯(Bill Lawrence)表示，“浆料的实际回收 利用量是一个取决于硅片尺寸、硅片 厚度、线锯设置以及浆料中硅含量的 一个变量。对于含油浆料来说，用来 清洗硅片及设备的清洗剂的回收率一 般都在90%以上。因此，浆料的回收率 越高，可重复使用材料的回收量就越 大，从而使得花费在新材料上的资金 就越少。” 尼克·萨诺(Nick Sarno)对于浆 料回收利用的益处见解颇丰，其在赛 维LDK太阳能任职期间曾针对回收再 利用作出许多战略举动。萨诺先生表

图6：CRS Reprocessing Services经典 泥浆回收系统。

18

w w w. pv -te c h .o rg

示，赛维LDK太阳能拥有一套3万吨级 的回收系统(见图7)，同时，公司还将 在硅片年产能增至2GW以上的同时加 装一套相似的设备。萨诺先生反复强 调，成本规避是公司发展中的重点因 素，而材料的回收利用可有效降低新 材料的使用量。 而意料之中的是，CRS公司的劳 伦斯先生仍旧反对硅片制造商试图设 计修建厂内回收系统的趋势，坚持认 为将这一业务交由拥有全套供应设备 及管理系统的第三方专家是最好的选 择。 “在厂内自行进行浆料回收利用 的缺点就是其沙砾和硅片承载器的回 收率均十分低。通常，此种操作仅能 回收极小的一部分硅片承载器，仅为 20-30%左右。此外，自行进行回收利 用将很有可能影响到产品质量的监控 以及公司的产能，并进而导致产品规 格下降、实验室验证工具受限和停产 时间延长等问题。” “就长远来看，可提高沙砾和硅 片承载器的最佳再处理解决方案，能 够将一套10线锯设备的月均总体运营 成本降低27.5至37.5万美元。上述这 些因素，再加上那些对于业务发展不 是十分总要的资金来源，可大幅减少 公司期望借由厂内回收再利用系统而 获得的潜在业绩。” 总体来看，当谈及浆料再处理 时，企业通常会面临四个选择：自行 处理、在第三方场地进行处理、在生 产现场进行处理以及不进行浆料再处 理。 劳伦斯先生指出了目前在欧洲盛 行的在第三方场地进行再处理工艺的 一些相关费用问题。他认为，使用在 第三方场地进行再处理所产生的服务 费用其实是十分具有相对竞争力的， 甚至可能低于常见的现场再处理工艺 的花销。但是，在第三方场地进行再

处理工艺也却确实会产生一项其他方 面的不可忽视的费用。 其中最主要的一块就是运费，而 这其中分为货运和航运两部分。运费 的高低在很大程度上取决于浆料的多 少以及运输距离的远近。在某些情况 下，这笔费用是十分庞大的。通常情 况下运费的费率为在境内每千克3美 分，而跨境运输为每千克40美分。如 果一名客户每月要进行800吨的浆料再 处理，那么其每月在此项目上的花费 就高达20万美元以上。同时，在第三 方场地进行浆料再处理也要求制造商 拥有大量的替代浆料以保证自身的生 产业务。 上述所提及的许多问题已在如半 导体制造业这种平行产业内引起极大 的关注，在半导体制造业内，大多企 业都采用现场回收再利用的方式，并 受惠于其所带来的其他益处。 就产品质量监控来说，现场浆料 再处理对于制造商来说具有完全的操 作透明性，并可即时获得质量过关的 浆料。同时由于浆料从未离开过生产 区域，整个过程均在封闭的作业流程 中进行，因此可以避免外界污染进入 生产流程的风险。此外，专业人士在 进行的现场测试时也能够轻松便捷地 进行各种调节，以获取质量稳定优良 的浆料，进而有助于在保证低废料率 的同时获得硅片的高产量。 在此，笔者再次强调，产品质 量是整个产业的发展核心，即使在祈 求简单的成本分析的过程中也不应忽 视。优化产量和提高质量可更有效地 节约成本。 ·三万吨级浆料回收系统已于2009 年9月投产使用 ·公司将于2010年中期新增一套三 万吨级设备

图7：赛维LDK太阳能的厂内泥浆回 收系统。

想了解更多吗？ 请登录PV-Tech.org在线杂志库: www.pv-tech.org/journal_archive

晶体硅太阳能电池发射 极形成的若干方法

Jan Bultman, Ilkay Cesar, Bart Geerligs, Yuji Komatsu & Wim Sinke, ECN Solar Energy, Petten, 荷兰

此技术文章节选自第八期印刷版《Photovoltaics International》季刊

摘要： 晶硅太阳能电池的核心组成部分是发射极，即p-n结。大部分硅电池的生产通过采用单步高温扩散工艺来实现对硅 晶格的掺杂。本文详细介绍了该扩散工艺的特性及工艺过程中所涉及到的变量，试图通过对这些变量的研究对发射极 工艺进行优化。此外，还介绍了其他制作发射极的技术，以期实现太阳能电池效率的最大化。

发射极的基本特性及优质发 射极的性能需求 理想情况下 太阳能电池工作的基本原理是光 吸收后产生电子-空穴对，电子和空 穴扩散或漂移到相应的电荷选择界面 处，再在界面处分开成为正电荷或负 电荷（即收集过程，见图1）。电荷的 收集导致界面两端形成电势差，也就 是我们所说的电池电压。当外接电路 时，电荷将会流动形成通路，从而产 生电流。 就电池的实际应用而言，最重要 的参数莫过于电池的输出功率，也就 是电池的电压和电流。电子-空穴可能 会被束缚成对（产生激子），也可能 会被单独分开，这种特性会导致不同 的器件，设计要求大大不同。在晶体 硅太阳能电池中，通常电子-空穴对 是不受束缚的，这意味着所产生的电 子和空穴能够独立地移动。用作电荷 分离的标准界面是p-n同质结。这里 的p和n分别指的是p型和n型掺杂，而 “同质”指的是掺杂发生在相同的半 导体材料上，即晶体硅。这种结构实 际上就是一个双极硅二极管。除同质

结外，还有一种被称为异质结的电荷 界面，在这种结中，两种不同的半导 体材料，例如晶体硅和非晶硅被结合 在一起。 与电子-空穴对产生相反的过程是 复合。当硅吸收光并产生过量的电子空穴对时，即会偏离热平衡方程，这 将会导致净复合的发生。电子和空穴 在到达p-n结之前即发生复合，阻止了 电子-空穴地分离，进而导致输出电流 的下降。复合的另一个不利影响是会 降低输出电压，本文将在随后详细探 讨此效应。因此，太阳能电池的制造 和设计的精髓就在于如何减少复合， 同时加大电子、空穴被分离和收集的 机率。 晶硅太阳能电池最常用的结构是 平面型二极管（如图2）。在这种结构 中，中等掺杂浓度的p型或n型硅片前 表面沉积覆盖有一层薄薄的、极性相 反、重掺杂的n+或者p+型硅，这一重 掺杂区域通常被称为发射极，而中等 掺杂浓度的材料，即硅片，则被称为 基极。“发射极”这个词将在更详细 的解释p-n结特性时得以理解。发射极 是指在（无光条件下）“发射”（注 入）大量载流子的区域。在晶体管学

前电极 n型硅

-+ -+

+ + 背电极

图1 晶体硅太阳能电池在光照下的截面示意图

p-n结 p型硅

中，“发射极”、“基极”和“集电 极”指的是几个不同的器件区域。 目前，大多数商业化晶硅太阳能 电池采用p型硅作为基底硅片，但以 n型硅作为基底硅片的太阳能电池现 在也受到了人们越来越多的关注。究 其原因还是由于：与p型硅相比，n型 硅具有硼氧复合结构且对杂质的敏感 度较低，它的光致引发衰退现象较不 明显。此外，整个制造过程对于p-n 结是否要置于电池前表面以及是否应 采用平面结构等并无具体要求。一个 最典型的相关例子就是背接触式太阳 能电池。在这种电池结构中，p-n结 是以重掺杂区的形式存在于器件背 面。此种电池又被称为交叉背接触式 (Interdigitated Back Contact，简 称IBC)太阳能电池，由SunPower公司 研发并已经实现商业化生产。 在目前的标准工艺中，发射极 是通过高温扩散工艺将n型杂质(磷) 掺杂至含硼的p型硅表面。由于所掺 杂的磷的浓度通常要比基底自身所 含硼的浓度要高得多，因此硅片表 面将由p型转换为n型，并最终形成 p-n结。也就是说，此区域是由“相 互补偿”的材料组成的。在上述过 程中，当p型和n型活性掺杂浓度相等 时，被称为冶金结。在冶金结的两端 存在着耗尽层（又称空间电荷区）。 耗尽层是载流子的消耗区域，只能包 含具有固定电荷的、离子化的掺杂原 子，因此也被称作空间电荷区。在 结的n型一侧，空间电荷是正的，而 在p型一侧，则是负的。需要注意的 是，整个空间电荷为零：从而，整 个器件仍旧保持其电中性。冶金结 两侧空间电荷区的宽度也因此取决 于各自的掺杂浓度。对于中等掺杂 浓度（掺杂浓度约为10 16 cm -3 ）基底 上的重掺杂型发射极（10 19 cm -3 ）， 基底一侧整个耗尽层的厚度大约为 1µm。在空间电荷区内会形成电场，这 一电场力可抵消由于结两侧存在的高 浓度差而产生的载流子扩散力，从而 Photov ol tai c s I nter nati on al

19

冶金结

方程（4） 其中，Wsc表示空间电荷区域的总 厚度，电池的电流贡献区域则仅限于 结一侧的扩散长度的范围内。 第三个决定太阳能电池效率的参 数是填充因子FF： 发射极 (n型)

空间电 荷区

基极 (p型)

方程（5）

图2 太阳能电池p-n结示意图，图中表明冶金结周围包括电中性的发射极、基极和空 间电荷区。从图中可以看出，与蓝光（短波长）相比，红光（长波长）可在距硅片表 面更深的区域产生电子-空穴对

在结的内部建立起平衡。需要指出的 是在耗尽层外部的材料，是不存在电 场的。 为了更好地理解太阳能电池发射 极的设计和制造需求，有必要来理解 一下理想太阳能电池在光照下的反应 方程式[1]：

方程（1） 此处电压函数J(V)是太阳能电池 的输出电流密度，J 0 是二极管饱和电 流密度（也称为暗电流密度），q是 电量，k是波尔兹曼常数，T是绝对温 度，J 1 是光致电流密度（通常相当于 短路电流密度Jsc）。 p-n结二极管的饱和电流密度J0的 计算方程式：

方程（2） 此处D和L分别是p型硅（多为基 底）中少数电子(e)和n型硅（多为发 射极）中空穴(h)的扩散系数和扩散长 度。本征载流子浓度n i ，在给定温度 下是一个常数。Na是p型硅中受体掺杂 浓度，而Nd是n型硅中施体掺杂浓度。 值得指出的是，在热平衡方程中，ni2 = [e]·N a = [h]·N d 。这将少数载 流子浓度与多数载流子浓度之间建立 起了联系，并进一步与掺杂浓度建立 起关系，此时，所有的掺杂原子都具 有活性，或者贡献电子，或者贡献空 穴。 扩散长度L（如：少数载流子在 复合前移动的平均距离等）取决于扩 20

w w w. pv -te c h .o rg

散系数D和少数载流子寿命τ，其相 互间关系为 。L和τ还受复合 程度的影响。电子性能较好的材料通 常具有较长的少子寿命和较长的扩散 长度，但此处的“长”是一个相对的 概念，必须根据器件的尺寸来定义。 通常有三种复合机制在发挥 作用：辐射复合（真正的光逆吸 收）、缺陷能级引发复合（也称为 Shockley-Read-Hall, SRH复合）和 俄歇复合。缺陷能级的产生是由晶体 材料的体内缺陷或者表面材料不纯所 引起的，晶体硅由于具有间接带隙， 这种间接的带结构性质使得光吸收和 辐射复合的作用相对较弱，那么缺陷 导致复合和俄歇复合就成为复合的主 要机制。凭经验而言，缺陷导致复合 限制了工业用中等掺杂浓度的硅片质 量，而俄歇复合则主要发生在重掺杂 工艺过程中（即便是在高质量、高纯 度和低缺陷的硅片中，俄歇复合也会 发生）。表面复合则取决于缺陷。 Jo,base和Jo,emitter的大小及其相对 重要性也可能会随着实际器件和材料 的参数的变化而大幅改变。对于太阳 能电池的优化而言，基极和发射极等 零部件都需要被纳入考虑范围内。 由方程（1）中可以导出，电 池开路电压V oc（V@J=0）可由下式确 定：

方程（3） V oc值 最 大 时 ， 所 对 应 的 J 0值 最 小，同时也就是短路电流密度J sc最大 （假定与J1相等）。 在极简化模型中，假定材料特性 和生成速率G是常数，短路电流密度可 由下式导出：

此处Jmp和Vmp分别表示最大输出功 率所对应的电流和电压。对于理想二 极管而言，FF只是Voc的一个函数，但 是在实际应用中，FF还受限于其他因 素，本文将会在后面的部分对此进行 解释。 非理想二极管特性：表面和电阻效应 方程（1）、（2）和（6）显示 了未考虑表面效应的理想二极管的特 性，即如图2中，当Wemitter和Wbase都是 无限大时的情形。但考虑到实际使用 中，发射极和基底的有限尺寸以及存 在的表面复合，则需要用一个更具体 的、考虑了表面效应的方程来表示：

方程（6） 此处Fp和Fn是以下参数的函数： S——表面复合速率（S和少数载 流子浓度[e]或[h]的乘积即为载流子 在表面复合的通量） ——厚度W与长度L的比率（L， 结的“可控跨度”，表征了受表面特 性影响的区域的典型厚度）。如果 >>1，表面质量对器件的影响可以忽 略；如果 <<1，器件质量则主要受表 面特性的影响。 ——扩散速率。等效于表面复 合速率。如果太阳能电池的基极或者 发射极区域是“无限厚”的话，复合 完全取决于体内复合。在基极或者发 射极区域稍薄一些的条件下，当表面 复合速率被设定为 时，新结构中的 复合等于“无限厚”时结构的复合速 率。 因此，表面质量达到S<< 时， 在光的吸收得到充分保障的条件下， 可通过使用较薄的或电子穿透性更好 （即 值较小）的硅片或者（发射极或 者背表面电场）薄层，使器件性能得 以提升。 需要指出的是，方程（1）并没有 考虑到串联电阻(Rse)和并联电阻(Rsh) 的影响，也没有包含空间电荷区的复

方程7

合效应，而这些都将会导致引发非理 想二极管的特性。这里可用电流词汇 Jon来表示，其理想因子n≈2。注意二 极管特性中的横向均匀性，例如串联 电阻的局部变化以及少数载流子寿命 也可能会导致（明显的）理想因子n>1 等。综合考虑这些因素，则电流-电压 特性可由下面的方程7来表述。 R se和 R sh即 所 谓 的 集 总 参 数 ， 该参数综合考虑了器件不同部位 的特性。这显然只是一个对于较 精确的2D或3D器件模型的一个近 似值。该表达式表明，用于控制 二极管电流的、实际施加在器件 中结上的电压，可能会低于器件 的 终 端 电 压 （ 如 应 用 电 压 V a等 ） 。 这将会引起填充因子的损失，进而会 影响电池转换效率。 在实际加工状况良好的器件中， 并联电阻R sh的作用常可被忽略，而串 联电阻R se则可通过优化来最大限度的 实现器件性能。串联电阻的大小与电 流在器件不同部位的传导性有关，其 中与发射极有关的包括以下几种（见 图1）： • 载流子从发射极到接触区的水平 传输（发射极体电阻，是发射极 电阻在深度上的积分） • 金属-硅接触界面的传输（接触电 阻） • 前金属电极的传输 实际发射极： 如上文所提到的，发射极通常是 由掺杂原子扩散到硅片表面形成的。 栅指

到目前为止，掺杂技术还不能达到在 整个扩散层都进行均匀掺杂的程度。 在无限源扩散中，扩散后杂质浓度分 布为余误差函数分布。而在有限源扩 散中，则呈理想的高斯函数分布。因 此，方程2中的第二段等式可以看作 是深度的函数，而方程6则在形式上 更为复杂一些。很明显，这只能用 PC-1D[3]等数值模拟工具来进行研 究。除了这些微小的修改，我们还需 要考虑一些其它的因素。对于发射极 等掺杂浓度梯度（dN d /d x ）较高的区 域，即便是在外部的耗尽层，仍然可 形成电场ε [3]。

方程（8） 尽管这一电场强度与耗尽层电场 相比弱得多，但它却可以通过对载流 子施加较小的漂移力，而帮助于少数 载流子扩散入结（从而避免其因扩散 至表面而被复合）。 发射极中的最佳掺杂浓度分布 （峰浓度、形状和深度）可以通过多 参数分析来进行评估。并且，前表面 金属电极下被遮挡的发射极区域和位 于金属栅指之间的发射极区域的最佳 掺杂浓度分布是不同的。这就会引出 所谓的选择性发射极的研究。在选择 性发射极中，两种区域的掺杂浓度分 布是不同的，金属电极下区域无需吸 收光线以产生载流子，但该区域需要 实现金属与硅的低电阻接触（比如： 倒金字塔

双层减反射膜

薄氧化层 p型硅

背表面 图3 钝化发射极背部局域扩散（PERL）电池

氧化层

使多数载流子在穿越界面时不发生明 显损失）。 而且，金属电极下区域需要能够 阻止少数载流子在欧姆接触中的过量 复合，因为欧姆接触所产生的复合通 常具有较高的复合速率。这通常会导 致一个相对较深的掺杂浓度分布，这 种分布通常会表现为很高的表面浓度 和明显的掺杂梯度。高表面掺杂浓度 也确保了低电阻接触通道的形成[7]， 同时掺杂浓度、掺杂深度和掺杂梯度 的协同作用减少了表面复合，对于方 程6中的 >1，其中Leff是发射极中的 有效扩散长度，值得注意的是，由于 俄歇复合作用的增加，L会随着掺杂浓 度的降低而降低。 金属栅指间的发射极在设计时需 要综合考虑到以下几个方面的因素： • 载流子的有效收集 发射极中由于光吸收而产生的载 流子（决定了短波长光的内量子 效率） • （多数）载流子的低损耗横向传输 指由载流子被收集的位置向距离 其最近的金属电极传输的过程 （这种向发射极体电阻的传输不 仅与金属栅指间的距离有关，同 时又取决于为减少过量遮光损失 而制作的最小栅指宽度）。 • 最大输出电压，见方程（3） 乍看之下，这似乎是在要求尽量 提高掺杂浓度，但当考虑到扩散 长度会随着掺杂浓度的降低而减 少时，就会发现掺杂浓度存在一 个最佳值而不应是一味追求最大 值。这在很大程度上取决于发射 极表面钝化的可行性。 在实际操作中，对所涉及参数的 优化（考虑到工艺设置的边界条件） 会导致栅指间的掺杂浓度分布明显不 同于位于金属电极下方的区域。当表 面涂层的钝化效果做的很好时，少数 载流子的表面复合速率能够被有效降 低。从而可将发射极中的整体掺杂浓 度降到最低值，也就是横向电流传输 时应满足的低电阻流失条件。与金属 电极下的区域相反，活性发射极区域 可被描述为 <1，此时所产生的载流 子可得到有效的收集，但同时也减小 了方程6的等式右边部分，并因此提高 了输出电压（见方程3）。这一论点可 被进一步强化为：对于载流子的收集 而言，发射极越薄越好。发射极（耗 尽层和中等掺杂的基底）区域下的载 流子收集效率通常高于高掺杂的发射 极本身。 然而需要强调的是，对于工业 电池的具体设计而言，需要考虑表面 复合速率的实际最小值，这可以通过 调制其函数——表面掺杂浓度来实现 （参见方程6中S与 的关系）。 Photov ol tai c s I nter nati on al

21

新南威尔士大学马丁•格林教授发 明的、开创了世界纪录的、效率高达 25%的电池技术即是成功的采用了选择 性发射极的概念（见图3）。这种钝化 发射极背部局域扩散（PERL）电池甚 至对选择性发射极技术进行了进一步 改进——即使发射极接触区域宽度窄 于顶部金属栅指宽度。这一精妙的设 计为硅-金属界面的表面复合和接触电 阻损失提供了更好的平衡。 重掺杂、死层和吸杂 标准扩散工艺采用无限源杂质扩 散。在这种工艺中，发射极表面区杂 质含量等同于该杂质在所采用温度下 的固溶度。当温度在850-950℃之间 时，磷的固溶度大约是3*10 20 cm -3 (即 约½%的硅原子被一个磷原子所取代)。 理想情况下，活性载流子（电子）浓 度应等同于磷原子浓度。在这种载流 子浓度极高的情况下，俄歇复合作用 更为显著，此时载流子寿命和扩散长 度都很短，且磷原子无法均匀分布， 所以其浓度可能更高，从而扭曲了硅 晶格结构，并进一步导致了缺陷复合 的发生。在这种情况下，并不是所有 的掺杂原子都是活性的，化学态的磷 浓度可能会高于临界电子活性浓度。 这样的表面层即会表现出极短的少子 寿命，因此也被称为“死层”。“死 层”的出现将会严重破坏电池性能， 特别是在活性、未遮蔽的发射极区

域。如果“死层”的形成不可避免的 话，还可以通过化学刻蚀的方法或者 通过在较厚扩散层中进行磷原子杂质 再分布的推进扩散法，在其形成后将 其移除。或者，也可以通过减少杂质 源浓度或者采用通过阻挡层扩散的方 法来避免“死层”的形成。 在特定情况下，较高的发射极 掺杂浓度还可用于提升电池性能。重 掺杂（晶格变形）层可在吸杂的过程 中作为汇集杂质的“杂质槽”。高温 工艺（如扩散等）下，电池基底中少 子寿命较短的杂质活性可得以提升。 如果重掺杂发射极区中的杂质有效溶 解度高于太阳能电池基底中的杂质有 效溶解度，大多数杂质将存留在发射 极中(即被“吸附”)。如果这些杂质 在发射极中产生的负面影响小于其在 基极中产生的影响，电池性能将会得 到提升。鉴于发射极复合通常是由俄 歇复合来决定的（与杂质和缺陷辅助 效应相反），并考虑到发射极对饱和 电流密度J o 的贡献可能要小于基极对 其的贡献，上述情况的发生也不无可 能。

参考文献 [1]

[2]

[3]

[4]

[5]

Green, M.M.A. 1992, Solar Cells: Operating Principles, Technology and System Applications, University of New South Wales, Kensington, Australia. van der Heide, A.S.H, Schönecker, A., Bultman, J.H. & Sinke, W.C. 2005, Progress in Photovoltaics: Research and Applications, Vol.13, 3.

PC-1D [available online at http:// w w w . p v. u n s w . e d u . a u / l i n k s / products/pc1d.asp].

Muller, R.S. & Kamins, T.I. 2003, Device Electronics for Integrated Circuits, 3rd Edition, John Wiley & Sons, New York.

Zhao, J., Wang, A., & Green, M.A. 1999, “24.5% Efficiency Silicon PERT Cells on MCZ Substrates and 24.7% Efficiency PERL Cells on FZ Substrates”, Prog. Photovolt: Res. Appl. Vol. 7, pp. 471-474.

鸣谢 澳洲国立大学的安德烈·奎瓦斯 教授(Prof. Andres Cuevas)和罗布· 埃利曼教授(Prof. Rob Elliman)为 本文中提及的掺杂运用内容提供了非 常有益的建议，在此我们表示感谢。

想了解更多吗？ 请登录PV-Tech.org在线杂志库: www.pv-tech.org/journal_archive

PV-Tech 博客 减本已成为“老大难”问题 ——对全球上网电价补贴调整趋势的调查研究 随着时间的脚步迈入了2010年的金秋，关于上网电价补贴下 调的消息也不再是使人们惊讶的新闻。继近期法国政府宣布削减 规模在30m2以上太阳能项目的补贴额后， 彭博新能源财经频道 随后的报道称，捷克共和国、安大略省和英国等地将可能纷纷效 仿此举。 全球各国都面临着光伏产品价格下跌的问题，同时，完成抑 制CO2排放量的任务也迫在眉睫。太阳能设备安装价格的下跌，激 发了公众对此技术兴趣，从而，政府所发放的补贴款就通过电费 转嫁到了消费者身上。 伦敦彭博新能源财经频道的一名分析师表示：“太阳能光伏 面板生产价格的大幅下降迫使各国政府在消费者通过缴纳其可再 生能源电费而为部分补贴款买单时，不得不降低所发放的清洁能 源补贴额。” “当初在制定补贴价格时，各政府并没有意识到组件和相关 系统会在如此短的时间内有这么大幅的下降。”彭博新能源财经 频道首席太阳能分析师珍妮·蔡司(Jenny Chase)表示，“我们 预计英国、捷克共和国和安大略省的政府也将对光伏产业进行相 应的政策调整。”

现有太阳能项目的发展态势将给消费者带来“巨额”的电力费 用。 据菲舍尔先生表示，为太阳能发电提供政策支持的整个框 架，在一开始定价的时候就不甚理智。“这给我们上了惨痛的一 课。”费舍尔先生还指出，“因此，在所有亟待批准的政策中， 该政策(削减上网电价补贴额)是重中之重。” 在今年下半年，捷克共和国下议院的议员已通过了下调刺激 政策的决议。拥有200个席位的捷克议会，以169比1的票数通过 了下调针对太阳能发电的巨额补贴的决议。此举直接引发了举国 上下对未来电价急剧上涨以及电网稳定性方面不切实际的幻想。 该决议若想成为正式的法律条款，仍需获得上议院的同意以 及总统的签字授权。该法案一旦通过，将使得能源监管办公室得 以针对2011年新增的安装设备下调上网电价补贴额。与安省的政 策相似，该下调举动的可行性极大。

感谢First Solar公司友情提供图片

英国

安大略省 截止至2010年7月1日，在加拿大安大略省微型发电上网电价 补贴政策下的光伏系统已达到1.6万套。其中大多数设备为地面 支架型系统，因此，安省电力局决定针对此类型中小于10KW的设 备进行上网电价补贴额下调，以减轻纳税人的负担。 许多证据都表明，目前安省之所以有逾一万份太阳能安装申 请被搁置，就是由该政策调整所直接导致的。能源基建部长布拉 德·杜吉德(Brad Duguid)表示，若没有进行此次调整，安省的 纳税人将在未来的20多年里为此支付约10亿加元(约合65.1亿人 民币)的资金。“任由这种状况肆虐下去是一种十分不负责任的 做法。” “安省电力局认为新的价格体系是公平合理的，并能更准 确地反映出地面支架系统的成本和保持设备长期稳定方面的花 销。”安省电力局首席执行官科林·安德森(Colin Andersen)表 示，“此举可保证系统能够完成其既定目标，同时向电力生产商 和纳税人提供更合理的价值体系。” 由于安省所存在的问题较为具体，因此补贴额下调政策的实 施问题不大。据悉，于八月二日前提交的申请仍将按照原有政策 享受补贴。

捷克共和国 捷克共和国首相杨·菲舍尔(Jan Fischer)早在今年三月份 就曾呼吁对可再生能源的补贴额进行削减。首相表示，对整个国 家来说，下调上网电价补贴是“首要”任务，如若不尽快执行，

英国对此措施的态度不甚明了。英国能源部长查尔斯·亨得 利(Charles Hendry)在最近接受英国《每日邮报》采访时指出， 于2010年四月一日起实施的极为慷慨的上网电价补贴很可能将在 下一轮政府支出审查中被削减。 “我们所面临的情况是，我们很清楚都有哪些方面将从中获 取商业利润，但在如何分配资金以及法律制定方面的做法仍不明 了。”亨得利先生表示，“我们一定要保证将钱用在刀刃上。” 独立智囊团Policy Exchange公司近日公布了一项名为 Greener, Cheaper(更环保、更便宜)的调研报告。该报告着重研 究了降低成本的重要性，并就在不对经济或社会福利进行无谓牺 牲的前提下实现减缓环境变化的目标寻求可行性方案，同时强调 了上网电价补贴政策对于促进英国可再生能源领域发展上的重要 性，以及该政策给消费者带来的经济负担。报告称：“针对微型 设备所发放的资金，以及其他控制环境变化的手段，均将会被转 嫁到电费花销上。同时，公众对于电费上涨的接受程度是极为有 限的。” 然而，必须考虑在内的一个因素是，所有的政府开支，无 论是直接花销还是如同转嫁电费这种间接花销，包括上网电价的 补贴政策，其通过与否都取决于下一轮的审议决定。与安省和捷 克情况不同的是，英国的上网电价补贴政策刚刚实行不久， 其 目的是想促进英国太阳能产业的发展，而不是减缓其增长速率， 有鉴于此，并没有确凿的证据表明此次补贴额下调政策能否付诸 实践。 正如上文所提到的，目前光伏电力的成本正在逐步下降，因 此，为将相关设备安装量保持在一定的水平而在全球范围内下调 上网电价补贴额，确是一项必须的手段，同时也可为由能源供应 商身上转嫁到纳税人身上的成本设置了上限。但是，这种情况具 体什么时候会发生仍不明了，特别是对于英国这种刚刚实施相关 政策以促进产业增长的国家来说，情况更是如此。立法者必须谨 记的是，下调上网电价补贴额很有可能使得国内可再生能源产业 的发展停滞不前，进而使自己处于无法完成限制二氧化碳排放目 标的被动位置。 本文选编自PV-Tech.org网站博客。

作者：Emma Hughes，Photovoltaics International杂志记者/ 专题编辑，同时为PV-Tech.org网站撰写新闻及博客。

Photov ol tai c s I nter nati on al

23

为全球顶尖太阳能制造商提供明智的解 决方案：更快捷，更简便，成本更低!

“经综合评估市场上 具竞争力的解决方案 后，我们认为Eyelit 是制造业公司可升级 软件的首选” Solyndra

“相较于之前在其他光伏制 造企业中使用过的MES解决 方案， Eyelit的能力和灵活 性使我印象深刻” Ascent太阳能技术公司 2010年3月5日

2010 年6月3日

“我们对各种不同的解决 方案进行仔细和严格的筛 选，最终认为Eyelit解决 方案是我们全新晶圆厂的 最佳选择” Deutsche Solar AG 2010年2月25日

Eyelit 为我们提供了扩大生产 需要的所有功能，并在关键总 体拥有成本上占明显优势” CaliSolar

2009年2月12日

“我们需要一套既能满足 我们目前成长的需要，又 能在将来帮助我们实现规 模化生产的解决方案” InnovaLight 2080年7月11日

www.e yelit.com A4.indd 1

30/09/2010 17:06:52

New Products AEG Power Solutions’ ‘Thyrobox M’

I

Oerlikon Leybold Vacuum’s DRYVAC

I

Bluestar Silicones’ ‘CAF 530’ adhesive range

Photovoltaics International Volume 06 - 2010

Methods of emitter formation for c-Si solar cells ECN Solar Energy examines the options

Cost reduction & productivity improvement strategies for multicrystalline wafering processes The PV-Tech Blog:

Investigating likely feed-in tariff changes worldwide

www.pv-tech.org

Turn Our Silver into Your Gold

Only if you have a flare for profits Heraeus silver paste is proven to increase the performance, value and output of your solar cells.

Our front side pastes offer higher efficiencies and greater aspect ratios to create superior electrical performance.

This increased output means more value to your customers and lower cost per watt for you.

Something the competition can’t deliver.

Our back side bus bar silver pastes are lead-free and offer industry-leading adhesion and coverage.

And we can customize formulations to meet your exact requirements. At production plants and application labs in China, Taiwan, Europe and North America, our dedicated personnel provide unmatched local technical and sales support to respond quickly to your business needs.

Call us, we can help you quickly realize the benefits of working with Heraeus.

Heraeus Materials Technology LLC Photovoltaic Group West Conshohocken, PA 19428 USA +1 610-825-6050 www.pvsilverpaste.com techservice.hcd@heraeus.com

Learn more. Visit our website or contact us today.

A4.indd 1

29/09/2010 12:32:04

Introduction Published by: Solar Media Ltd. Trans-World House, 100 City Road London EC1Y 2BP, UK Tel: +44 (0) 207 871 0123 Fax: +44 (0) 207 871 0101 E-mail: info@pv-tech.org Web: www.pv-tech.org Publisher: David Owen Managing Editor: Síle Mc Mahon Senior Editor, North America: Tom Cheyney Senior News Editor: Mark Osborne News & Features Editor: Emma Hughes News Editor, North America: Syanne Olson Production Manager: Tina Davidian Design: Daniel Brown Commissioning Editor: Adam Morrison Sales Manager: Neill Wightman Account Managers: Adam Morrison, Graham Davie, Daniel Ryder, Gary Kakoullis, David Evans, Nick Richardson & James Park Marketing Manager: Joy-Fleur Brettschneider

Global solar demand is soaring, and all the major PV manufacturers up and down the supply chain have either completely sold out of capacity (even after aggressive capacity expansions) or are running at very high factory utilization rates. European end-markets are continuing to fuel industry growth, particularly Germany, Italy, France and the Czech Republic.

However, Asia-based PV suppliers are reaping the rewards of a booming 2010 global market. According to recent second-quarter figures released by market research firm Solarbuzz, six Chinese manufacturers accounted for 55% of solar cell shipments, up from 43% a year ago.

Key developments have included the companies’ ability to add capacity at aggressive rates, much faster than their western peers, while maintaining the ability to reduce manufacturing costs and improve overall c-Si cell efficiencies.

In this edition, two key articles characterise why Asia-based PV manufacturers are riding the current demand cycle. A round-table discussion on low-cost wafering technologies and strategies highlights the attention to detail and cost reduction strategies that are dictating successful wafering operations today. A technical paper from ECN discusses the hot topic of solar cell selective emitters, which is a key process innovation being adopted by leading high-volume cell manufacturers – many of whom are now Asia-based – to push towards 20% conversion efficiencies.

With market research firms rushing to raise installation figures past 16GW for 2010, Asia is rapidly becoming the major manufacturing hub meeting the global demand.

While every effort has been made to ensure the accuracy of the contents of this supplement, the publisher will accept no responsibility for any errors, or opinion expressed, or omissions, or for any loss or damage, consequential or otherwise, suffered as a result of any material here published.

Mark Osborne Senior News Editor Photovoltaics International/PV-Tech.org

Cover image shows a PV system at the Aspen Mountain Ski Resort. Image courtesy of NREL. Credit: Aspen Skiing Company.

Printed by Shing-Hsiang Printing Co. Photovoltaics International Lite Volume 6, 2010 ISSN: 1757-1197 The entire contents of this publication are protected by copyright, full details of which are available from the publisher. All rights reserved.

No part of this publication may be reproduced, stored in a retrieval system or transmitted in any form or by any means – electronic, mechanical, photocopying, recording or otherwise – without the prior permission of the copyright owner.

Contents 2 News 12 Products

Image courtesy of Sputnik Engineering /SolarMax

Copyright: Sonel d.o.o.

14 Cost reduction and productivity improvement strategies for multicrystalline wafering processes Photovoltaics International

Photovoltaics International Lite PV Taiwan edition has been produced for exclusive distribution at PV Taiwan 2010, providing access to a handful of technical papers, product reviews and opinion pieces, and serves as an introduction to the full subscription version, which features more than 20 technical papers from leading industry figures each and every quarter.

19 Methods of emitter formation for crystalline silicon solar cells ECN Solar Energy 23 The PV-Tech Blog: Cuts conundrum: Investigating likely feed-in tariff changes worldwide Emma Hughes

Annual subscription available for US$199.

For more information call Nick at +44 (0) 207 871 0123 or email nrichardson@pv-tech.org

P hot ovol t ai cs Int er nat i on al

1

News World first: SoloPower is awarded UL certification for flexible CIGS modules In a breakthrough achievement in flexible solar photovoltaics technology, SoloPower has received Underwriters Laboratory (UL) certification for its lightweight flexible CIGS modules, a first-ever accomplishment in the thin-film industry. This is the first UL-certified product in a range of high-power flexible modules being introduced initially to the European and North American markets.

The UL certification was granted following rigorous testing at an independent laboratory where the flexible thin-film modules were tested to UL 1703, the standard for safety for PV module manufacturing. SoloPower also claims to have conducted extensive internal testing exceeding the safety, quality, and reliability standards established by these tests.

“The certification of SoloPower’s flexible CIGS module is an important step toward the realization of lightweight, high-power, flexible solar modules with potential to expand the roof-top solar market and reduce balance of system costs. It is an important milestone for the industry. I feel very gratified to see, after a 30-year career in Thin Film CIGS PV at NREL, the technology become mature,” said Dr. Rommel Noufi, principal scientist of the NREL. SoloPower’s range of flexible photovoltaic module products include the SFX1 module (80Wp, 0.3m x 2.9m, 2.3kg/5lbs.), the SFX2 module (170Wp, 0.3m x 5.8m, 3.6kg/8 lbs.), and the SFX3 module (260Wp, 0.9m x 2.9m, 6kg/13lbs.).

Thin-Film News

DuPont’s PV8600 series to be used for Sharp’s solar cell modules DuPont Photovoltaic Solutions has disclosed that its new ionomer encapsulant sheet, known as Himilan ES in Japan and PV 8600 series elsewhere, will be used by Sharp in its thin-film solar cell modules. The encapsulant was made for amorphous silicon and other thin-film module technologies and has recently entered the commercial market. “ I o n o m e r- b a s e d s h e e t s a r e f u n d a m e n t a l l y m o re re s i s t a n t t o moisture intrusion, current leakage and discoloring,” said Steve Cluff, global business director for DuPont Encapsulants. “Their use in glass-glass module designs is well-established, but adhesion to polymer-based backsheets, such as DuPont Tedlar and PET, required modifications that we’ve introduced in the new DuPont PV8600 Series sheets.”

DuPont advised that with the exception of Japan, the encapsulant will be available for customer evaluation later this year.

Sunvalley Solar to use a-Si thin-film modules from TianWei SolarFilms U.S.-based PV installer Sunvalley S o l a r h a s a d d e d m o d u l e s f ro m TianWei SolarFilms to its list of product offerings. Sunvalley Solar has supply agreements for Canadian Solar’s poly and mono crystalline panels as well as for CEEG. TianWei SolarFilms will supply an unspecified amount of a-Si thin-film modules to Sunvalley Solar, which are now CEC (California Energy Commission) certified.

TianWei SolarFilms’ CTO Mai Yaohua presents the company’s production process to a guest.

TSMC to sell own-branded CIGS thin-film modules as first 200MW fab starts construction Sharp’s solar modules use DuPont Photovoltaic Solutions’ PV 8600 ionomer encapsulant sheet.

2

www. pv - t e c h .o rg

Moving away from its ‘pure-play’ foundry business model to which it has vehemently adhered in the s e m i c o n d u c t o r i n d u s t r y, Ta i w a n

Installation of SoloPower’s lightweight flexible CIGS modules.

Semiconductor Manufacturing Company (TSMC) will directly sell copper-indiumgallium (di)selenide thin-film modules to the global market from its first 200MW plant, which broke ground recently. An initial US$258 million is being invested to build a Thin Film Solar R&D Center and production plant, using technology from Stion. The thin-film start-up is partnering with TSMC, which will be its manufacturing and technology development partner.

TSMC also announced plans to add a second phase to the facility and expand CIGS production to more than 700MW, employing around 2,000 staff at the facility in Taichung’s Central Taiwan Science Park, home to its leading-edge semiconductor and newly-announced foray into the LED market.

First-phase equipment move-in was said to be planned for the second quarter of 2011, with plans to achieve initial volume production of 200MW per year in 2012. No timelines were given for the larger, second-phase capacity expansion.

“TSMC has always been committed to technology leadership, and our solar business will be no different,” commented Dr. Rick Tsai, TSMC president of new businesses. “The research performed at this R&D center will help us achieve our goal of offering a leading thin-film solution and the production at this fab, drawing on TSMC’s wealth of manufacturing knowhow, will pave the way for us to become a top provider of solar PV modules.”

The first CIGS facility was said to be 110,000m2 in size with a production area of 78,000m2.

TSMC chairman and CEO Dr. Morris Chang said, “Construction of this solar R&D center and fab, along with our Fab

SunEdison signs MOU with Korean province to develop 400MW of solar PV power plants

TSMC’s headquarters.

15 Gigafab next to it, means Taichung’s Central Taiwan Science Park will become home to much of TSMC’s most advanced and innovative production.”

Power Generation News

Power-One to build 1GW inverter production plant in Phoenix Power-One has chosen Phoenix, AZ, as the site for its new manufacturing facility.

The plant will produce the company’s PV and wind inverters, including single(2-6kW) and three-phase string inverters, as well as NEMA 3R 250kW, 300kW, and 400kW central inverters.

The company said that manufacturing at the Arizona facility will begin in October and that annual inverter production capacity at the site will reach 1GW by mid-2011. The new factory is expected to support more than 350 new jobs in the state and throughout the country.

Power-One said it has recently expanded its worldwide production capacity with the initiation of Canadian manufacturing and the expansion of its European plant, which will result in more than 4GW of inverter capacity coming online by year’s end.

The company also plans to launch Chinese production in the fourth quarter with products specifically designed for the requirements of Asian markets. The estimated cost of the new plant was not disclosed.

Power-One’s Aurora inverter system.

4

www. pv - t e c h .o rg

In a potential deal that would dwarf the scale of the company’s previous development projects, MEMC subsidiary SunEdison has signed a memorandum of understanding with the Gyeongsangnam-do provincial government for the establishment of 400MW of PV power plants in the southeastern Korean province. The nonbinding MOU, which is subject to negotiation and completion of definitive agreements, could lead to the utilization of public land and building rooftops to develop and install the PV systems.

The provincial government said it will support SunEdison in securing the proper land or building areas and in completing the authorization and permission processes. The projects will be completed by the end of 2013. No financial details of the proposed deal were disclosed.

SunEdison is expected to reinvest certain amounts of profit received from these projects for the development of the local economy and will try to encourage other companies located in Gyeongsangnam-do to participate in the construction and installation of the solar power plants.

D u r i n g M E M C ’s m o s t r e c e n t quarterly earnings call in late July, the parent company said that SunEdison has 111.5MW of project work under construction that would be completed by the end of 2010. The Korean deal, if approved, would lead to projects that would be equivalent to approximately double the amount of SunEdison’s entire expected portfolio by year’s end. It would also be one of the largest collections of PV installations in the east Asian country.

the known advantages of a distributed inverter architecture without the module power and reliability limitations that have traditionally held back this class of product.”

“While this was a significant milestone in our ‘go to market’ evolution, we are planning a global product launch, so additional regional certifications are expected over the next several weeks, including UL certification for the U.S market,” he concluded.

SunEdison wins contract for 14.5MW plant on DavisMonthan Air Force Base, Arizona Adding to the company’s ever-expanding portfolio of large-scale installations, SunEdison has won the contract to provide a 14.5MW ground-mounted solar power installation at the DavisMonthan Air Force Base (DMAFB) in Tucson, Arizona. The agreement will see SunEdison look after the financing, design, construction, operation and maintenance of the ~130 acre site, which, upon completion, will provide as much as 35% of DMAFB’s energy needs.

SunEdison has also entered into a long-term ground lease agreement with DMAFB for the use of their land, and in return, will provide the DMAFB with electricity supply at a fixed rate.

“With our strong financing capabilities and our track record of timely largescale deployments, SunEdison is able to provide smart solar solutions with no upfront capital costs and long-term predictable energy pricing,” stated Jaime A. Smith, vice president of Sales for SunEdison. “SunEdison is proud to be awarded this contract and looks forward to working with DMAFB to the completion and activation of the solar power plant.”

Direct Grid gets CE, VDE certification for utility-grade solar PV microinverter Direct Grid Technologies has received CE and VDE0126-1-1 certification for its DGM-S460 Series utility-grade grid-tie photovoltaic microinverter designed for large commercial and solar farm installations. The inverter will be distributed in Europe initially through OEM partner One Network Energies in France, according to Direct Grid, with additional partner agreements near finalization and set to be announced in the near future.

“We are very excited about receiving CE and VDE certification for the DGMS460,” said Direct Grid’s president, Frank Cooper. “This is an industry first, being that we have positioned ourselves in the 500W class in European utility grid deployment. Customers can realize

Air show at Davis-Monthan Air Force Base.

Gigawatt scale: CEC approves Solar Millennium’s planned 1GW Blythe CSP power plant The California Energy Commission has given its unanimous approval to Solar Millennium to build and operate the Blythe Solar Power Project in Riverside County, a planned 1GW concentrated solar power (CSP) plant that would be the largest solar generating station in the world. Construction on the first pair of four planned 250MW parabolic-trough

THE 2010 THIN-FILM ANNUAL Your definitive Hardback Guide to the World of Thin Films in 2010

ORDER ONLINE TODAY FOR €35! (+ €5 delivery)

www.pv-tech.org/shop

The Rise of Thin-Film Solar Technology is the definitive handbook and resource for professionals working or interested in this exciting sector of the solar photovoltaics industry. This volume compiles authoritative technical and market-related articles written by over 70 industry experts from the pages and databases of the Photovoltaics International quarterly and the pv-tech.org website.

A comprehensive contact directory and product reviews offer indispensable resources for anyone determined to understand and navigate the thin-film PV segment.

Sign up online for your free Thin-Film Insights Newsletter!

Photovoltaics International

lite

JobsinPV.com Recruitment

PV-tech.org Daily News

SolarLeaders Te l e v i s i o n

PV DIRECTORY PV DIRECTORY

Contact: Photovoltaics International | Trans-World House | 100 City Road | London | EC1Y 2BP | United Kingdom | info@pv-tech.org | www.pv-tech.org | +44 (0) 207 871 0123

ECN receives IEC 61215 certification for MWT module

Solar Millennium’s Andasol 1, the first parabolic trough power plant in Europe.

CSP sections is scheduled to begin by the end of the year.

Although the CEC approval was critical, the company must still secure a “Record of Decision” approving the project’s “Right of Way Grant” from the Federal Bureau of Land Management, which is expected to occur this autumn.

The Solar Trust of America project development subsidiary said it is also actively pursuing completion of financing with the U.S. Department of Energy Loan Guarantee Program for the initial 500MW phase of the project.

The project will generate approximately 2500 jobs during the construction period and create more than 200 permanent jobs once the 1GW facility is fully operational, according to the company.

Once completed, the dry-cooled CSP plant will produce enough energy annually to supply more than 300,000 homes.

Solar Trust says it has nine utility-scale solar thermal projects in advanced stages of development in California and Nevada.

The Blythe plant is the third major CSP farm to gain CEC approval recently, following the recent go-aheads given to NextEra Energy’s proposed 250MW B e a c o n p ro j e c t a n d t h e 2 5 0 M W Abengoa Mojave Solar Project. The commission is expected to rule on several other CSP projects before the end of the year.

ECN has achieved the IEC 61215 industry standard certification for its metal wrap through (MWT) module technology. The certification was awarded in collaboration with an undisclosed solar photovoltaics company as part of a joint research framework agreement.

Paul Wyers, manager of ECN Solar Energy said, “After years of development we were not only confident about the performance, but also about the robustness of our back-contact modules. The back-contact module production method is beyond state-ofthe-art. We use materials that are new for crystalline silicon modules, such as conductive adhesives and conductive back-sheets. We are very happy that our MWT module technology has been awarded the certificate by TÜV-SÜD and we are confident that modules with this technology will be sold in the near future.”

“From the moment we started with promoting our equipment in 2008, there has been strong interest in this unique technology”, said Bram Verschoor, commercial director of Eurotron. “The MWT module certification allows us to sell turn-key production lines. A single Eurotron line is capable of assembling 150MWp of MWT modules per year with a footprint comparable with a 20MWp tabber stringer.”

IEC 61215 is a recognized industry standard that verifies that the modules have passed the required series of tests and are suitable for long-term operation in a range of climates.

CNPV to supply 20MW of modules for China-based projects

6

www. pv - t e c h .o rg

field of 500m 2 , with equipment such as five climate chambers and two sun simulators. This makes it one of the most advanced PV testing laboratories in the entire South Asian economic area.

“Our investment programme for the solar industry aims to place our services within easy reach of companies in and for all booming markets and to offer them large test capacities. India cannot be neglected here.

In addition, all our customers can call upon the decades of expertise gained by our now 180 experts around the world to test and certify systems, modules and components,” said Friedrich Hecker, CEO of TÜV Rheinland.

“All our laboratories are working closely together and contribute to our global PV business. We are delighted about the opening of the new Indian facility which will enable us to cover the expanding demand and expected growth in the Asian region and continue to offer tailor made solutions that suit the needs of our customers,” says Stefan Kiehn, head of the PV testing facilities at TÜV Rheinland Japan.

Fab and Facilities News

Solarwatt doubles module capacity to 400MW; new line produces module every 28 seconds

PV Modules News

Module manufacturer CNPV is providing Linyi Juhuang New Energy Technology Development Company with 20MW of PV modules through 2012 for a series of 1MW projects in Shandong Province, China. Initially, 1MW will be supplied in October 2010 and the remaining 19MW during 2011 and 2012.

“The first grid connected system is on target for connection using our premium solar module range,” said B. Veerraju Chaudary, CNPV’s COO, CTO, and member of the board. “This bespoke system will add a further 25% yield increase on our premium range’s market leading capability. It is an exciting time for us to be further involved in Shandong Province’s homegrown solar electricity generation.”

TÜV Rheinland India’s Bangalore headquarters.

ECN’s 17.0% conversion efficiency MWT module.

TÜV Rheinland opens solar testing laboratory in Bangalore TÜV Rheinland has now officially opened its seventh solar modules and systems testing laboratory, based in Electronics City in the Indian city of Bangalore. TÜV invested €2 million in the new solar test centre, which will offer services to India’s burgeoning solar industry.

The test centre includes 2,000m 2 of space, including an outside test

Touted as the ‘world’s most modern module manufacturing facility,’ Dresdenbased Solarwatt has officially opened its new production lines that double capacity to 400MW and produce a solar module every 28 seconds, according to the company. A new 13,000m2 logistics centre has also been built, which includes a 260kWp solar power plant.

The total investment was €35 million.

“By establishing the world’s most modern module manufacturing facility a n d b e g i n n i n g p ro d u c t i o n t o d a y Solarwatt is meeting the challenges of coming years,” said CEO Frank Schneider.

As is the norm for many Westernbased manufacturers, the key to module manufacturing competitiveness is to have highly automated plants. According to Solarwatt, the new line employs 29 industrial robots from Kuka Systems, which include its Robo Frame, Robo Trimm, and Robo Load solutions.

PV DIRECTORY The leading online sourcing directory for the solar industry – find the perfect fit for your supply chain.

YOU CAN BE THAT MISSING PIECE! PV Directory offers you the fastest and most cost-effective way to generate new business opportunities in the solar supply chain. It offers an affordable alternative or addition to your advertising and PR campaign by providing company- and product-specific information to business professionals in the solar industry. The online directory also attracts businesses that are looking to enter the high growth solar industry.

Benefits at a Glance f Put an always up-to-date summary of your company in front of keyy industry decision makers who are looking to source suppliers. f Showcase your specific products and solutions in words and images. f Make your services searchable against multiple product and geographical criteria. f Exploit the Photovoltaics International branding and tive image to maximise authoritative the exposure for your company. f Boost traffic to your corporate website and improve your SEO efforts.

If you’re not in front of them, then who is?

Reach over 150,000 visitors and 329,000 pageviews per month as part of www.PV-tech.org!

Benefit from the most authoritative PV website worldwide and channel real business to your website!

Please contact David Evans devans@pv-tech.org or Nick Richardson nrichardson@pv-tech.org or alternatively, dial: +44 (0)20 7871 0123

pv_directory.indd 1

22/09/2010 11:01:27

EU laws do not allow such ‘local content’ rules; now that Japan is raising the issue of the pricing guarantees offered by Ontario’s FiT system actually constitute subsidies that are not allowed under international trade law.

Dr. Ulrich Link, COO of Solarwatt presents the company’s new production line to Stanislav Tillich, Premier of Saxony.

A total of 140 new jobs were created working a three-shift system, the company said.

Danfoss announces expansion plans for its Nordborg headquarters As a result of what Danfoss Solar is calling a period of very strong sales, the company has announced plans to expand its production and logistics capacities by moving to larger facilities at the company’s Nordborg, Denmark headquarters. Danfoss states that it has outgrown its supply chain facilities in Gråsten and Sønderborg, Denmark and with this move will be able to not only increase their 2011 capacity to 3.5GW, but have the ability to double the assembly size in a few months.

Production at the new location is expected to start in the first quarter of 2011, with the move finalized by the end of next year. In addition, Danfoss will be expanding its China and U.S. facilities to facilitate the demand for solar inverters.

“Expanding our activities in Denmark, the U.S. and China will enable us to achieve the growth and earning targets for Danfoss Power Electronics that have been set out in the new strategy for the Danfoss Group,” says Troels Petersen, president of Danfoss.

SMA Solar raises revenue guidance a second time: expects PV market to top 17GW in 2010 The global PV market is set to soar higher than previously expected, according to solar inverter market leader, SMA Solar Technology. As a result, the company has raised its revenue guidance for the year, the second time it has done so in 2010. SMA Solar now expects the global photovoltaics market to reach approximately 17GW of new installations in 2010 and with more than 40% of the inverter market, is guiding revenue of between €1.7 and €1.9 billion. Previously, SMA Solar had forecasted the PV market would reach 14GW and company revenue of €1.5 to €1.8 billion.

SMA Solar said that Germany would remain the largest single PV-market with new installations reaching 8GW. South European markets and the U.S are also fuelling growth. However, forecasting growth for 2011 is proving more difficult, according to the company.

On the one hand, it expects PV global growth of 20% in 2011 but uncertainties surrounding various FiT changes and possible negative impacts on the market could in another scenario see the PV market decline 10% next year.

Key countries that could dictate growth levels for 2011 were identified as the U.S. and Italy, which could become very important growth markets next year.

Market Watch News

Japan protests to WTO about Ontario’s locally produced product quotas for FiT Ontario’s locally produced product quotas that are required to benefit from its lucrative feed-in tariff have upset the Japanese, according to Reuters, which has led to an official complaint to the World Trade Organisation (WTO).

The two sides are set to meet but the likely outcome is to move proceedings to a WTO panel that would make a judgement about the legality of Ontario’s FiT requirements.

There have been numerous announcements from PV manufacturers throughout the supply chain that are setting up facilities in Ontario to comply with the rules as they see the local market highly attractive.

8

www. pv - t e c h .o rg

owned California utilities to purchase electricity from renewable energy systems between one and 20MW in size.

“California has robust policies for developing large, utility-scale solar power plants and for putting smaller systems on homes and businesses, but there is a clear gap in the middle. The CPUC proposal is designed to unlock that missing piece, providing an additional opportunity for solar market and job growth and for quickly bringing massive new amounts of clean energy to the state,” said Adam Browning, executive director of Vote Solar, who will work with CPUC to implement these changes.

“Solar policy should provide the foundations for long-term market growth by providing a transparent process, a level playing field, and a reliable market opportunity,” said Kevin Fox, of the law firm Keyes & Fox, which represents IREC, another advocate of the initiative.

“This program achieves those larger policy goals through an innovative pricing mechanism that also protects California ratepayers and overcomes the legal challenges that have hindered widespread feed-in tariff development in the U.S.” The CPUC proposal establishes a 1GW pilot program for power from eligible mid-sized renewable energy systems. The program requires California’s three largest investor-owned utilities to hold biannual competitive auctions into which renewable developers can bid. Utilities must award contracts starting with the lowest cost viable project and moving up in price until the MW requirement is reached for that round.

The program will use standard terms and conditions to lower transactional costs and provide the contractual transparency needed for effective financing. Development security and relatively short project development timelines ensure project viability. The commission can act to finalize and adopt the program in as soon as 30 days.

Cell Processing News

SMA Solar Technology’s Sunny Boy inverter.

California will drive midsized solar projects with new incentive program The California Public Utilities Commission (CPUC) has issued a proposed decision to launch a new renewable incentive program with the aim of driving the uptake of mid-sized renewable energy development. This next-generation feedin tariff program will require investor-

JA Solar nabs supply agreements for 500MW throughout 2011 JA Solar Holdings has secured multiple s u p p l y a g re e m e n t s w i t h s e v e r a l undisclosed customers, which will see the company delivering over 500MW of mono-crystalline and multi-crystalline solar cells throughout 2011. The signed contracts will see deliveries begin in January 2011, continue through December 2011 and have prepayments for next year’s committed solar cell deliveries. In addition, JA Solar released that its year-end capacity for 2010 will be 1.8GW with actual productive capacity at 1.4GW for the end of the second quarter.

JOIN TODAY! THE GLOBAL NETWORK FOR FORWARD THINKING BUILDING SPECIALISTS!

www.design-buildsolar.com is the first and most authoritative resource for architects, engineers and installers ready and willing to integrate pv into their overall design projects. The website provides the latest news, technical articles, videos and product information for all stakeholders while supplying a platform for architects and designers to display their latest projects through case studies, share best practices and request product information from key suppliers.

THE STRONGEST COMMUNITY BIPV HAS EVER SEEN

“We are seeing very strong demand for our technologically advanced, highquality products from our diverse global customer base,” said Dr. Peng Fang, CEO of JA Solar. “Our customers view JA Solar as a reliable supplier partner, and rely on JA to provide them with their 2011 product requirements. We are pleased to have secured these new supply contracts for next year delivery, which further improves our 2011 visibility and reflects our continued leadership in supplying high-quality solar products to the industry.”

Comtec Solar to deliver over 600MW monocrystalline solar wafers in 2011 Comtec Solar Systems, together with its subsidiaries, has signed new wafer supply framework agreements, with price subject to negotiation, to provide major customers Gintech Energy, Jetion Solar, CHINT Group and Neosolar Power with a total of approximately 200MW in monocrystalline solar wafers. These orders are thus added to the existing agreements between Comtec and China Sunergy, JA Solar, Suntech and Canadian Solar.

Under the terms of the contracts, Comtec will supply each of the companies with approximately 50MW of monocrystalline wafers from January 2011 to December 2011.

Company: China Sunergy JA Solar Holdings Suntech Power Holdings Canadian Solar Gintech Energy Jetion Solar Holdings CHINT Group Neosolar Power

100MW 150MW 150MW 50MW 50MW 50MW 50MW 50MW

Total:

650MW

Comtec’s wafer supply deals for 2011.

LDK Solar completes first production line of solar cells at Xinyu City facility LDK Solar has successfully completed the installation and trial runs of the first production line of solar cells in its newly installed manufacturing facility in Xinyu City. The solar cell manufacturing line has a present annualized capacity of 60MW; however this figure is expected to reach 120MW by the end of the third quarter of 2010.

“We are very pleased to announce the commencement of our solar cell production,” said Xiaofeng Peng, 10

www. pv - t e c h .o rg

chairman and CEO of LDK Solar. “Once again our team has accomplished a rapid production ramp of a new facility. This is an important milestone as we continue to execute our integrated growth strategy. Our objectives of in-house cell production are to produce approximately 50% of our module manufacturing needs, thereby reducing the cost of our modules and attaining a stable supply of cells. Our newly established solar cell manufacturing facility in Xinyu City advances this strategy.”

Materials News

Silicon Genesis introduces PolyMax system for production of solar wafers Silicon Genesis (SiGen) has made a move towards what it deems as a replacement of the wire saw process with the production of solar wafers using its PolyMax system. SiGen has produced 85μm-thick, 156mm2 kerf-free monocrystalline silicon wafers. Kerf is a material that is translated into saw dust and used in the sawing process.

“We believe the benefit of using kerffree wafers will allow the PV industry to reach unsubsidized grid parity. The start up of our high volume manufacturing system is a key step towards achieving this goal,” said Francois Henley, CEO of SiGen.

CRS executes long-term contract with Nexolon Company C R S R e p ro c e s s i n g S e r v i c e s a n d Nexolon Company have endorsed a multi-year contract in which CRS will supply on-site reprocessing services at its production facilities. Nexolon is continuing towards its goal to reach a 1.5GW capacity by the end of next year.

“Nexolon is excited about the prospect of partnering with CRS. This company has an excellent reputation for increasing recovery rates while delivering cost savings to their business partners,” said Sung Joon Kim, director of strategy planning department of Nexolon. “Their program will deliver a consistent slurry quality with a variety of carriers while minimizing staining issues, and that’s exactly what we need.”

The plant is located in Iksan, South Korea, and will use CRS’s turnkey solution for reprocessing slurry from the production of solar wafers. The signed contract has an agreed term of five years with a four-phase build out.

It is predicted that at the plant’s peak production, the slurry yield will be up to 36,000MT per year.

Yingli Solar signs fiveyear polysilicon supply agreement with OCI Chemical Yingli Solar has signed a five-year polysilicon supply agreement with OCI Chemical, the leading chemicals producer in Korea. Under the terms of the agreement, OCI has agreed to supply polysilicon with a total value of approximately US$442 million to Yingli Green Energy from 2011 to 2015.

Yingli recently posted strong revenue, gross margin, and net income numbers in the second quarter of 2010, with sequential increases shown across the board. The vertically integrated Chinese solar manufacturer said it experienced a substantial increase in PV module shipments during the period ended June 30 and still expects to hit between 950MW and 1GW in total panel shipments for 2010.

OCI Chemical’s manufacturing plant.

CRS Reprocessing Services’ slurry recycling system.

Tokuyama’s planned polysilicon plant in Samalaju Industrial Park, Malaysia gets go-ahead Although site selection had been undertaken back in November, 2008, Tokuyama has now approved the construction of a 6,000MT polysilicon plant in the heavy industry based Samalaju Industrial Park in Malaysia. The plant will commence construction in early 2011 and start operations in the spring of 2013, according to the company. Plant costs were said to be ¥65 billion and utility costs of ¥15 billion.

The polysilicon plant will be dedicated to supply the solar industry. Previously, Tokuyama planned to start operations in 2012 and have a plant with 3,000MT capacity to supply both solar and semiconductor industries.

Special News Feature Heliene starts moduling operations at Ontario plant, but exports dominate order book

Image courtesy of Heliene.

No sooner had the “local content” provision of Ontario’s solar power feedin tariff been announced than a steady stream of PV module, inverter, and other balance-of-system players said they would build manufacturing sites in the province to take advantage of the lucrative incentives. The fun really begins in January 2011, when the FiT rules will stipulate increased local content and significantly ratchet up the clamor for homegrown panels.

In a new study released by ClearSky Advisors, the market research firm says that demand for “bankable” panels will exceed the estimated 386MW in production capacity due to come online in Ontario next year, which might trigger some project delays and premium pricing. Yet Heliene, one of those numerous companies getting into the locally-made module game, actually has a majority of its order book going to customers outside the province.

I first spoke with Martin Pochtaruk, president of the Sault-Ste. Marie-based venture, in late February, when he was waiting for the weather to warm up and construction of the 18,000-square-foot purpose-built facility to be completed. At the time, he told me of the plan to “get access to the finished building by the beginning of July,” with the equipment to be set up and tested at Heliene’s sister company in Spain and final installation

coming in early August. The plan then was for the first modules shipping out to Ontario first, and then to the U.S. Midwest, by the end of August, beginning of September.

As it turns out, they’ve come close to hitting that schedule, running only a month or so behind the initial timeline.

When I spoke to Pochtaruk in midSeptember, he said that most of the equipment was in the factory, with the last few pieces arriving soon and his first group of employees returning back to Ontario from two weeks of training in Spain.

Manufacturing for sales began in late September, with plans to be fully ramped, running four shifts 24/7 by January. But the initial IEC/UL-certified crystallinesilicon modules coming off the line will not be delivered to a customer in Ontario.

“That’s a funny thing,” he said. “Our first client is a company in Saskatchewan.

We have orders for export for October and November,” citing bookings from firms in the Midwest U.S. and Europe, specifically Italy and Greece. In fact, for the next quarter or so, less than half of the orders have come from Ontarian outfits, according to Heliene’s president.

The company isn’t targeting any specific market sector. Pochtaruk characterizes his downstream channel partners and customers as “a group of installers with orders of 200-250kW per order, small utility-scale clients ordering 3-5MW per order, and distributors with recurrent volumes of 200-300kW per month on a yearly contract,” adding that the factory is sold out for several months.

Because of improvements in manufacturing productivity, the moduling facility now boasts a 50MW nameplate c a p a c i t y, u p f ro m t h e o r i g i n a l l y announced 30MW. He attributed this improved production cycle to some changes in the automation approach in the lamination section of the factory floor, with each of the six laminators processing nearly twice as many panels because of the addition of load/unload tables and double-tasking robots.

Another enhancement involves a new handling tool, which takes the stillhot laminated panel up, elevator style, into a vertical tower for cooldown (thus increasing the cooling capacity without taking up more floor space) before transferring it to the flash tester, he said.

Capacity could be increased to 80MW by adding a third soldering machine, although “that will depend on market conditions. If we see the market calling

for the product, we’ll add the tool and increase capacity,” explained the pragmatic executive.

“We have a cautious startup ramp, because the equipment is new, the building is new, the employees are new, everything is new. It makes it very sparkly,” he laughed, “but you need to be cautious not to overstretch.”

There’s enough space for three full production lines, so the factory could eventually grow to 240MW, if all the capacity-enhancing improvements were implemented.

The plant will manufacture both multiand monocrystalline modules, split about 25:75 respectively between the two product lines. Heliene has contracted three mono-cSi cell suppliers for 2011 – Suniva, Arise, and Bosch – and one multi-cSi vendor, Sunways. Pochtaruk also indicated that Saint-Gobain supplies the glass for the panels, and STR provides its EVA encapsulant materials.

The production gear, except for the flash tester, has been made in-house by Heliene’s sister company in Spain.

Since the factory construction and equipment set-up have gone reasonably smooth, what has been the most challenging aspect of getting things rolling? The newness of the PV industry in Ontario, according to Pochtaruk.

“We’re not just starting up a new company, it’s a startup of the whole industry because solar in Ontario is new,” he pointed out. “No one knows what has to be done. Most of the developers you meet don’t know if they’re coming or going. Even with the volume of [project] approvals that the Ontario Power Authority has made, there’s not much going on yet. Everything else, like financing, engineering, still needs to get done. It will be happening, otherwise I wouldn’t be here. But it’s taking time.”

Until things get going, Heliene continues to quote prospective customers in Europe, the States, and even a few in Ontario, biding its time until the local market finds its footing and the potential of that very promising FiT starts to be fulfilled.

This column is a revised version of a blog that originally appeared on PV-Tech.org.

Tom Cheyney is North American Editor for the Photovoltaics International journal and writes news and blogs for PV-Tech.org.

P hot ovol t ai cs Int er nat i on al

11

Products ISRA Vision has introduced an advanced optical multi-panel inspection system Applications: Inspection of laminates including incoming glass inspection, edge inspection, coating inspections (such as TCO) and scribing inspection.

Platform: ISRA Vision employs standard modules and standard components, engineered and manufactured in-house. This practice guarantees the highest level of customer flexibility. Systems are claimed to have a high acceptance and performance with safe and repeatable error recognition, unique real-time capabilities, as well as robustness and reliability.

Availability: Currently available.

ISRA Vision’s Powerscan system is an advanced optical multi-panel system for automated inspection of glass, surface and laminates for the photovoltaic industry.

Based on its standardized modules and software, ISRA has developed a special system setup to inspect thin-film solar modules right after lamination. The Powerscan system detects laminated bubbles, shrinkage and excessive PVB as well as edge defects and scratches on the glass itself. Based on its modular product structure in regard to the cameras, illumination and frames, the system can solve nearly every inspection task.

Additional features include state of-the-art interfaces, inspection of flexible batches and high-performance inspection tasks for thin-film solar products.

TwentyNinety’s MLC technology improves photovoltaic array performance Applications: Manages up to eight strings of Active Array modules.

Platform: The Active Combiner is a multi-channel device, able to manage up to eight strings of Active Array modules down to two outputs with full intelligent reverse current protection.

Availability: Currently available.

The ‘Active Array’ Module Level Control (MLC) technology offers a low cost wireless application for improved performance and increased safety for arrays of PV modules. The system enables the tracking of performance at PV module level, identifying problems to allow early intervention, optimising architecture and monitoring warranty data. It identifies and isolates a poor performing or shaded module, and thus improves string performance. Active Array also tackles fire and electrical safety concerns by providing remote data and control over every module. Intelligence at the module level gained wirelessly by the Active Array system provides a complete picture of the PV array, enabling the early diagnosis of potential problems.

Kemper offers dual-axis tracking system for large-scale power plants Applications: Large solar farm projects. Project specific individual constructions are possible.

Platform: Torsion-free, hot-dip galvanised steel construction that is suitable for all wind and snow loads adhering to DIN 1055. Extremely high wind protection position can be individually configured. The system has durable and industry tested drives. The central control unit can accommodate modular extensions such as a surge protector or the emergency power supply using batteries including an online evaluation of the charge states can be made.

Availability: Currently available.

12

www. pv - t e c h .o rg

Kemper’s ‘KemTrack’ 120 system is capable of handling a combined module area of 120m2. The construction is torsion-free, robust, durable and modular in construction. The KemTrack 120 dual-axis tracking system handles more modules compared to its other systems. A current yield of about 17kWp per tracker is claimed compared to 8.6kWp for its KemTrack 60 system.

The industrial drives used are “made in Germany,” maintenance free and extremely stable. The astronomical regulation of the two drives contains innovative functions and provides an increase of the current efficiency of up to 40%, according to the company.

Software algorithms prevent mutual shading with sun exposure angles.

Bluestar Silicones sealing and bonding offers elasticity over temperature range Applications: Sealing and bonding of PV junction boxes and general module applications, including thermal solar panels.

Platform: Bluestar Silicones has performed ageing tests according to IEC 61215 in order to check the bonding properties of the CAF adhesives on PVF and also on composite PVDF/PET composite materials for 1,000 hours at 85°C with 85% relative humidity, known as the damp heat test.

Availability: Currently available.

Bluestar Silicones’ ‘CAF 530’ adhesive range optimizes the longevity performance of solar modules by extending their resistance to weather e ro s i o n u n d e r c o n d i t i o n s w h e re performance must be constant for at least 25 years. Bluestar Silicones’ CAF 530 adhesive has a very high degree of elasticity, which resists and absorbs differential dilations of the assembled materials. The new range includes monoand bi-component silicone elastomers (RTV 1 & 2) designed for sticking and sealing the frame and junction box as well as bi-component products designed for the encapsulation of components in the junction box and of photovoltaic cells.

‘Thyrobox M’ power solution reduces energy consumption per 1kg of polysilicon Applications: Polycrystalline silicon deposition process.

Platform: System solutions from AEG are designed to interface with the electrical power grid and to offer power solutions for mission-critical applications in harsh environments, such as power plants, offshore oil rigs, chemical refineries, and utilityscale renewable energy plants.

Availability: Currently available.

AEG Power Solutions’ ‘Thyrobox M’ is said to be the most compact power solution, offering increased reliability and stability for the polycrystalline silicon deposition process. With advanced process monitoring capabilities, a 0.95 power factor and over 99% energy efficiency, the Thyrobox M reduces installation, commissioning and maintenance costs for the entire energy supply. Depending on the type of reactor, its footprint is about 25% smaller than older Thyrobox solutions and up to 70% smaller than competing products.

Warning system and troubleshooting features also minimize the risk of process interruptions in case of silicon rod cracks or melts. An integrated medium-voltage ignition enhances process stability.

DRYVAC pumps cover a variety of processing applications Applications: Wide-range of c-Si and thin-film manufacturing process applications.

Platform: DRYVAC pump systems and “-i” versions include housing, c a s t o r s , f re q u e n c y c o n v e r t e r, PLC and touch screen display as standard. Control and monitoring can be visualized via intuitive menu navigation, software and field bus.

Availability: Currently available.

DRYVAC, Oerlikon Leybold Vacuum’s range of dry vacuum pumps, have been designed and adapted to the special needs of coating applications and various industrial manufacturing processes. DRYVAC Sprinter and Enduro versions have optimized pumping speeds at all pressure ranges from 10-2 mbar up to atmosphere, making them ideal for fast pump-down times and suited for load-lock applications. Models from the DRYVAC Champion range provide the highest reliability in harsh process duties. The DRYVAC range comprises models: DRYVAC 650 S; DRYVAC 650 S-i; DRYVAC 5000 RS-I; plus multiple system combinations with roots blowers from the RUVAC series.

P hot ovol t ai cs Int er nat i on al

13

Cost reduction and productivity improvement strategies for multicrystalline wafering processes

Mark Osborne, Senior News Editor, Photovoltaics International

ABSTRACT Multicrystalline wafers are the workhorse of the PV industry, with approximately 60% of crystalline silicon solar cells made from the substrate. They offer cost advantages in the form of good conversion efficiencies, which should continue to improve as cell technology advances continue. However, wafer prices were acutely impacted by the fall in PV market demand in late 2008, which continued through most of 2009. With relatively high capital costs, continued pricing pressures and calls for greater quality and control, wafer producers are now set on a course that requires rigorous and sustainable production cost-reduction strategies to meet customer requirements. This paper focuses on strategies that can be adopted to address this need for tighter quality specifications that reduce manufacturing costs downstream and boost cell conversion efficiencies.

Introduction

demand creating a very tight supply environment, further capital expenditures will be necessary to meet customer demand. Traditionally in periods of strong demand, equipment lead times extend to over nine months as key production equipment such as furnaces and wire saws are slowly churned out. These tools are large in size and/or complex in construction making them inherently difficult to manufacture in high quantities at the short lead times.

“The erosion in pricing is bound to change the face of the solar industry,” noted Henning Wicht, Senior Director and Principal Analyst for Photovoltaic Systems at iSuppli. “The freefall of PV prices represents a permanent ratcheting down of price structures that will transform the industry into a more competitive marketplace.” iSuppli expects PV manufacturers to continue to focus on cost reductions in order to keep up with the price declines and to repair compressed profit margins experienced in 2009. When the costdown programs eventually catch up with the rate of price declines, an overall improvement in the profit picture can be expected, Wicht said.

This is of particular importance to polysilicon and wafer producers, who have experienced the most aggressive price declines, yet have significantly higher capital costs compared to others downstream, such as pure-play module manufacturers.

Challenges also exist for wafer producers. W ith the strong wafer

The role of polysilicon

Ccourtesy of iSuppli Corp.

Polysilicon, wafer and solar module prices all declined severely in 2009.

According to a recent report from iSuppli Corp., average crystalline module prices dropped by 37.8%, solar wafer prices fell by 50%, and polysilicon prices declined by 80%.

The market research firm expects further price declines in 2010, though not at the steep levels seen last year. In 2010, iSuppli is forecasting price declines for crystalline modules of 20%, solar wafers to decline by 18.2%, and polysilicon prices falling by a further 56.3% (see Fig. 1). Many industry observers now believe that with the continued growth in polysilicon production by both major and new entrant producers, declining prices along the supply chain are now set in stone for the industry.

Of course, this is a key prerequisite for an industry chasing the elusive grid parity (and beyond), which will benefit the players in becoming increasingly competitive with other renewable energy forms.

Figure 1. Average selling price forecast 2008-2012 in US$/MW.

14

www. pv - t e c h .o rg

Disconnects between points in the supply chain as seen recently in polysilicon shortages are in danger of transferring to the wafering sector. This results in greater pressure to boost productivity while new capacity is planned and eventually implemented.

As can be seen in Fig. 2, iSuppli has surveyed wafer production expansion plans of producers that show only a gradual expansion (left column) in wafer capacity based on actual planned capacity additions. The middle column shows the potential additional plans that could be undertaken but have yet to materialize. Only in 2012 and beyond will we see those plans add meaningful production capacity.

It could be argued that such a forecast only reinforces the pressures on wafer producers to optimize current capacity throughput, yield and overall productivity. New capacity additions could also benefit from new technology introductions that also reduce production costs. Those that successfully tackle these challenges will not only benefit from lower costs and improved margins, but also attract more customers

Ccourtesy of iSuppli Corp.

Figure 2. Global solar wafer production capacity plans (MW). downstream requiring lower prices to remain competitive. They will also gain advantage in a continuing overall price decline environment.

“Wafer producers have benefited from an overall improvement in material quality due to better availability of virgin polysilicon.”

There should be little doubt that p o l y s i l i c o n a n d s o l a r w a f e r s a re intrinsically linked. Wafer price declines have primarily been a result of the even greater decline in virgin polysilicon prices since mid-2008. Numerous market research reports point to continuing capacity expansions that should lead to further price declines. A recent report from Bernreuter Research estimates that global polysilicon production will reach 250,000MT in 2012, with approximately

80,000MT produced in China alone, making up about one-third of global production.

Johannes Bernreuter, founder of the research firm, told Photovoltaics International that polysilicon supply/ demand dynamics this year would translate into spot pricing in the range of US$45 to US$50/kg.

“In 2011, I expect that a significant volume in China will probably be produced at manufacturing costs below US$35/kg, and that the spot price will fall to this level – whether by the end of the year or earlier,” commented Bernreuter.

Wafer prices will therefore continue to decline, but as seen in 2009, prices fell much more than manufacturing costs, leading to significant margin declines. However, wafer producers have actually benefited from the plentiful supply of polysilicon in a different way.

According to Nick Sarno, the former VP of Manufacturing at LDK Solar and now currently consulting for the industry, wafer producers have benefited from an overall improvement in material

Figure 3. Bernreuter Research: Polysilicon supply demand pricing dynamics.

quality due to better availability of virgin polysilicon and the declining dependence for some using scrap silicon to complement ingot production.

“This translates into cost reductions throughout the wafer manufacturing process,” noted Sarno. “Less man hours on procurement of scrap silicon, sorting and storing has a small impact but better quality supply of virgin material produces better ingots, blocks and wafers, reducing scrap, consumables and delivering a better product all round.”

Furnaces: is size the solution?

In an effort to reduce ingot costs, there has been a gradual but growing trend towards larger ingot sizes. The larger the ingot, the more blocks and consequently wafers can be produced, resulting in increased overall yield increases (see Table 1). This therefore requires larger ingot growth furnaces. Much of the current ingot production is performed using Directional Solidification System (DSS) furnaces that cast multicrystalline ingots using 450kg furnaces for volume production. Developments continue on charge sizes above 500kg and 800kg, and there are efforts underway to increase ingot sizes to 1,000kg in the not-too-distant future.

In today’s market, a key limitation is the technological capability to manufacture crucibles of ever-increasing size. Common industry practice has been to tackle this obstacle by adopting larger furnaces but using multiple crucibles at a time within the furnace for optimization and greater throughput.

“From a capital expenditure point of view this strategy is a plus,” noted Sarno.

“You are future-proofing investments in anticipation of the larger crucibles being available, while not having additional investments for going bigger just because of crucible size limitations.”

Crucible manufacturers are attempting to address several challenges at once, answering increased demand for largersized crucibles – in particular 450kg products – while re-tooling for even bigger sizes in the future.

Ceradyne, a leading fused silica crucible manufacturer, is expanding manufacturing in the first half of 2010 in its existing facility, Ceradyne Tianjin Technical Ceramics (see Fig. 4), and has broken ground on a second facility, Ceradyne Tianjin Advanced Materials, which will be operational during early 2011, to meet growing demand.

A leading DSS furnace supplier is GT Solar, a company that has certainly seen the growing demand for larger systems in the last few years. During fiscal year 2009 and 2010, GT Solar sold approximately US$1 billion of capital equipment into the market to support the build-out of installed end-user capacity of approximately 8GW to 10GW.

P hot ovol t ai cs Int er nat i on al

15

Model Name

JZ-270

JZ-450

JZ-520

JZ-800

Initial Charge of Poly-silicon (kg)

270

450

520

800

Estimated Dimension of the Solidified Ingot (mm)

L&W

680

840

840

996

Thickness

250

270

315

345

Estimated Dimension of Cropped Blocks (mm)

L&W

156

156

156

156

Thickness

200

220

265

295

Estimated Weight of Cropped Ingot (kg)

181.4

311.8

375.7

602.2

Estimated Yield for Ingot (%)

67.2

69.3

72.3

75.3

Table 1. JYT large furnace benefit analysis.

Interestingly, the issue of larger furnace sizes was not the first point of conversation that Henry Chou, Product Marketing Manager, PV Equipment for GT Solar International wanted to mention. Rather, he noted that as the industry becomes more mature, GT Solar has brought to the table what he describes as the ‘value metric.’

Using simplified variables to highlight manufacturing costs, the company targets cost reductions via improved throughput of the tool coupled to higher yields and better overall ingot quality.

This then translates into less wastage from the ingot to the wafer sawing steps, also lowering costs.

“Being able to get better quality wafers as a result does indeed enable cell producers to achieve higher conversion efficiencies,” commented Chou. “What we look for is good kilograms produced per hour.”

“The two key metrics for cost reduction at the furnace are throughput and yield.”

The two key metrics for cost reduction at the furnace are throughput and yield, according to Chou. With respect to throughput, this relates to the charge size divided by the process time.

“Interestingly, if you start to add more mass to the charge, your process time will start to increase. Based on the process time, the quality of the segregation of impurities and growth of the crystals changes the optimum balance. So if you grow bigger ingots it doesn’t necessarily mean that you are

Figure 4. Crucibles at Ceradyne Tianjin Technical Ceramics.

16

www. pv - t e c h .o rg

going to get higher throughput, due to the growing cycle needing to be longer,” added Chou.

He also believes that ingot sizes above 600kg could lead to other equipment changes that actually add to the cost of production. Consensus on whether charge sizes above 450kg are inevitable would seem to be premature. The company has recently been tapped for its ‘DSS450’ furnaces and ancillary equipment and services to the value of US$200 million. The orders come from a range of wafer producers, including Tianwei New Energy, Phoenix Photovoltaic Technology, Yingli Green Energy, JA Solar and Sino-American Silicon Products (SAS).

In early May 2010, the company launched the ‘DSS450HP,’ a high performance ingot growth furnace with a thermally optimized, second-generation hot zone that is said to improve throughput while delivering industryleading crystal quality. Current users of the DSS450 and DSS240 models can migrate to the new DSS450HP with a field upgrade kit, which means that in the near term, wafer producers are regarding proven technology offerings as preferable to pushing the boundaries of even larger ingot/furnace sizes.

Slurry solutions Slurry is at the heart of wafer manufacturing. It is used with wire saws to cut the wafers out of silicon ingots, normally comprising an abrasive mixture of silicon carbide and ethylene glycol that erodes, rather than cuts, the ingot. Generally, a conversation centring on where key costs are located across the various wafering processes will lead straight to the topic of slurry.

As the slurry is fed along a guide wire, the abrasive silicon carbide in the slurry (or ‘grit’) cuts through the ingots to produce the wafers. Slurry should be regularly replaced and reprocessed in order to keep the number and quality of the wafers high while maintaining acceptable unit cost.

Although slurry is vital to the process, it also represents a significant proportion of the wafer production cost; slurry costs are often second only to the cost of the

polysilicon itself. Despite the relatively low price of US$3.50 per kilogram for new slurry, the overall cost can quickly escalate. In a scenario where 10 wire saws are being used in a process, the operating cost can be as high as US$16 million a year, not including disposal fees or labour charges.

For the vast majority of manufacturers, slurry reprocessing ranks in the top five contributors to operating cost. Optimizing this aspect of the business can have dramatic and lasting effects on wafer production costs and overall quality.

According to CRS Reprocessing Services, a leading slurry recycling firm, wafer manufacturers that are planning a slurry reprocessing program or starting from scratch in a greenfield operation can benefit from significant marginenhancing improvements. For a wafering operation running 10 wire saws with a volume of 385 metric tons per month, slurry reprocessing can result in a saving of anwhere in the region of US$8-10 million per year, net of recycling costs, utilities, and infrastructure investment.

“At CRS Reprocessing Services we are able recover between 80% and 90% of both the grit and the carrier,” noted Bill Lawrence, founder and president. “The actual amount of recovery is a function of the wafer size, wafer thickness, saw set-up and the silicon kerf loading in the used slurry. For oil-based slurries, the recovery of the cleaning solvents used to rinse wafers and equipment is typically greater than 90%. In short, the higher the recovery rate, the more material is effectively recovered and reused, and the less that must be spent on new material.”

The benefits of slurry recycling were echoed by Nick Sarno, who recalled a huge move toward recovery strategies during his time at LDK Solar. Sarno noted that LDK has a 30,000MT recycling system designed in-house (see Fig. 7), with a similar system being installed as the company ramped wafer production beyond 2GW per annum.

Sarno reiterated that cost avoidance was a key aspect as recycling significantly reduces the use of new material.

Perhaps not surprisingly, CRS’s L a w re n c e c a u t i o n e d a g a i n s t t h e tendency for wafer producers attempting to design in-house systems instead of using fully supplied and managed systems from expert third parties.

“The downside is that do-it-yourself options tend to have much lower recovery rates for both grit and for the carrier. These processes typically recover only a small portion, say, 20-30%, of the carrier. In addition, a do-it-yourself scenario may impact quality control and productivity resulting from lower specifications, limited lab verification tools and increased downtime.

“To put it in perspective, an optimal reprocessing solution that increased the recovery of both grit and carrier could

Figure 5. GT Solar’s DSS450HP ingot growth furnace. easily decrease the overall costs for a 10-wire-saw operation by $275,000 to $375,000 a month. These factors, along with the need to fund resources that are not central to the business, reduce the potential savings that are expected by taking the process in-house.”

In general, there are four options when it comes to reprocessing slurry: do-ityourself, off-site reprocessing, on-site reprocessing, or do no reprocessing at all.

Lawrence went on to point out some of the cost issues associated with offsite reprocessing, which is popular in Europe. He noted that the service fee for off-site reprocessing can be competitive, perhaps even lower than the common alternative of on-site reprocessing. There are a number of other costs, however, that factor into an off-site processing scenario that should be considered.

Chief among these are freight costs – both trucking and shipping, which can vary depending on the amount of slurry and transportation distance.

In some cases, the expense can be considerable. Rates range from $0.03 per kilogram within a country to up to $0.40 per kilogram around the world.

For a customer processing 800MT per month, this could total over $200,000 per month. Off-site reprocessing also requires that a manufacturer have large amounts of slurry ‘in play.’

Many of these issues have long pervaded in parallel industries such as semiconductor manufacturing, where most of the required recycling is done on site, bringing with it other benefits.

From a quality control standpoint, on-site reprocessing brings full transparency to the manufacturer, who gains the advantage of immediate and verifiable slurry. Since the slurry does not leave the facility, it is reprocessed in a closed loop that eliminates the risk of outside contaminants entering the production stream. On-site testing, conducted by experts who can easily and quickly adjust levels to achieve consistent and optimized slurry, helps ensure wafer yields are high with minimal waste.

Once again, the issue of quality is a key factor that should not be overlooked in the pursuit of straightforward cost analysis. Optimized yield and high quality products are often greater cost saving pursuits.

Chemical contribution Of course, slurry is not the only chemical used in wafer processing. Cleaning and texturizing steps use HF acids in baths.

Simply shopping around for lower cost bulk chemicals may prove effective in the short term, but once again, using the right chemistries for yield and wafer quality considerations could prove to be a greater contribution to cost reduction and optimization strategies.

Frank Haunert, Product Manager at BASF, commented that ‘solar-grade’ bulk chemicals such as HF acids that are not specific to semiconductor purity levels have been developed, which means that they can therefore be offered at lower prices, while maintaining the quality and consistency requirements that optimize processes. With respect to BASF’s ‘Seluris’ range of etching and texturizing chemicals, Haunert noted that they enable saw damage to be rectified and the wafer surfaces to be structured from ‘drop-in’ solutions, reducing quality inconsistencies and delivering a more homogenous wafer result.

With respect to slurry, Haunert noted they are developing slurries that have tackled issues such as wire pairing: “The typical wire saw can experience surface tensions that mean the wires cannot get through the slurry solution and start coming together creating one thin and one thicker wafer. This has a negative impact of yield and overall productivity.”

partly due to its simplicity relative to slurry-based processing. Although still in its infancy from an adoption perspective, the technology has some potentially compelling aspects.

“We believe we can drive down the wafering costs by 10 to 15% initially with diamond wire technology and further improvements later,” commented Peter Pauli, CEO of Meyer Burger Technology AG. “The important part of that is yield improvement. The industry as a whole only has an 80 to 90% total yield. If we only deliver a 5% yield gain this would be a dramatic improvement for the industry.”

However, concerns have been raised over the use of DWT, not only because of the inherent cost of diamond wire, but due to the physical difference of the resulting wafers compared to conventional wire/slurry-produced wafers.

Nick Sarno cautioned that diamond w i re l e a v e s s c o re m a r k s o n t h e surface of the wafer, making the wafers look physically different from non-DWT produced wafers. This has returned some negative feedback from customers, reporting that they are ‘a little afraid of using’ this type of wafer.

Meyer Burger’s CEO was well versed on this issue: “The sub-surface damage is actually less than with slurry,” Pauli pointed out. “The surface may look a little bit different, but changing to another technology takes time. We don’t expect a dramatic uptake of the technology right now.”

“Changing to another technology takes time” – the same can be said about changing perceptions regarding DWT. To that end, the company has made great strides in developing and importantly demonstrating the cost and yield benefits of DWT. Meyer Berger has invested positively in the complete system technology and infrastructure that would enable the industry to migrate to DWT at the high volume commercial level. Pauli noted that the actual diamond wire was only one aspect of the complete package needed to make adoption successful.

“The key is that DWT changes

Diamond wire solution

Figure 6. CRS Reprocessing Services’ typical slurry recycling system.

The new kid on the block for sawing bricks and wafers is diamond wire technology (DWT). A key bonus of diamond wires is that slurries are not required, therefore dispensing with the expensive material altogether. This of course means that there are no recycling costs nor any of the other costs associated with the use of slurry. DWT is also claimed to provide 2.5-3x the throughput compared to conventional wire/slurry combinations and is said to offer improved cutting accuracy, reducing kerf loss and providing stable processing

Figure 7. In-house slurry recycling system at LDK Solar.

P hot ovol t ai cs Int er nat i on al

17

Figure 8. Meyer Berger’s diamond wire comparison chart. wafering, internally and dramatically,” he added. “It gives us the opportunity to change and improve the whole line – simplifying and reducing steps and equipment needed in the value chain.

With overall complexity of the wafering process reduced, we have the ability to improve the overall yield and bring the industry into the industrialization phase.

This, to me, is the future.”

Thinking thin With the return of a plentiful supply of polysilicon, it would seem that there is less talk focused on migrating to thin wafers in order to reduce cost.

Although companies like LDK Solar have continued to migrate gradually to thinner wafers, there is less emphasis on taking wafers thinner than 170 microns, due to concerns at the cell processing level regarding higher rates of wafer breakage.

Although developments continue at the R&D level, production efforts remain focused on kerf loss.

Process optimization However, conventional wire/slurry technology is not stagnant. Significant efforts are ongoing to provide the yield and cost reductions required for the industry across the complete wafering process flow.

Herbert Arnold GmbH & Co. KG is a leading manufacturer of production systems for processing mono- and multicrystalline silicon bricks that incorporate cutting, grinding and polishing.

Speaking with Wolfgang Schürgers, Sales Director at Herbert Arnold, it was apparent that there are significant opportunities in regard to the optimization of equipment and processes that not only shorten cycle times but also provide greater understanding of the overall processes that in return improve productivity, quality and yield.

The introduction of advanced technology manufacturing systems improve process control and an integrated, process data analysis creates transparent and efficient production.

18

www. pv - t e c h .o rg

“What I can see is a lot of effort to optimize Gen5 ingots to produce 25 bricks of higher quality and improve the height of ingots by 300 to 400mm,” noted Schürgers. “I don’t see the trend towards the one-ton ingot. Kerf loss is still an important issue and we are able to provide a cut with the smallest kerf loss possible.”

Schürgers highlighted that cropping with blade saws rather than wire saws provides accurate straight cutting and lower cost. The development of high precision (1.5mm) blades with the optimization of the thickness of the smallest of blades significantly reduces kerf loss and boosts machine uptime, increasing throughput.

With the move towards larger wafer production lines in excess of 500MW capacity, there will be a growing trend towards fully automated lines, according to Schürgers.

Herbert Ar nold have calculated that automation of the entire process flow would generate at least a 10% overall productivity improvement. This is achieved by optimized 24-hour operations, reduced waiting times for tool availability and reduced machine idle time as a consequence. Importantly, highly optimized positioning of the brick and saw results in consistently improved yield.

Schürgers believes that high-volume facilities of 200MW and above that are still dependent on manual handling will have to deal with high error rates as a result of human operations, which have a considerable impact on the line’s ability to be fully optimized.

Optimization of the complete line was repeated as a necessity by Greg Knight, Director of Process Engineering Turnkey Services for GT Solar.

“Equipment is one piece in operating a factory,” remarked Knight. “How you utilize that facility is actually a bigger piece. We have done a lot of analysis on streamlining operations such as how we process material straight out of the furnace. There would seem to be a significant loss of revenue occurring in a lot of wafering facilities because they are

making the decision to cut on the marked brick lifetime lines… while this is a useful metric, using metrology to determine what is good and bad enables only the bad material to be turned into dust.”

According to Knight, better or less conventional cutting approaches can take yields to the 90% range when typically many wafer producers are operating in the sub-70% yield range.

Significant effort has been made with GT Solar’s turnkey lines to enable the feedback of data from sawing to ingots.

The multiple issues that can occur such as rejects and saw marks can be traced back to the specific furnace and correction procedures implemented in order to limit prolonged yield impacts.

Detailed monitoring of the operations enables tighter process tolerances that reduce reactive changes and so boosts productivity and reduces cost.

“Better or less conventional cutting approaches can take yields to the 90% range.”

Chemical usage optimization was also a factor covered by BASF’s Haunert.

He noted that efforts are ongoing to maximize the use of a given amount of material for the highest possible number of wafers. Aspects such as bath temperatures and flow rates are fine-tuned to enhance the chemical characteristics for processes such as wafer cleaning.

Conclusion As wafer producers recover from rapidly declining prices in 2009 and focus on a new wave of capacity expansions to meet strong demand in 2010, cost reduction continues to be a core focus.

The continuing abundance of polysilicon may well continue to place margin pressure on wafer prices, despite tight supply and a short-term recovery in wafer prices. Wafer producers that are vigilant in regard to cost and process optimization strategies will not only weather economic conditions better, but will likely flourish ahead of those competitors that fail to tackle these challenges in a coherent manner.

Figure 9. Fewer Si losses are possible through the optimal positioning of the block within the cropping saw.

Methods of emitter formation for crystalline silicon solar cells

Jan Bultman, Kay Cesar, Bart Geerligs, Yuji Komatsu & Wim Sinke, ECN Solar Energy, Petten, The Netherlands

ABSTRACT The emitter or p-n junction is the core of crystalline silicon solar cells. The vast majority of silicon cells are produced using a simple process of high temperature diffusion of dopants into the crystal lattice. This paper takes a closer look at the characteristics of this diffusion and possible variations in the process, and asks whether this step can lead to optimal emitters or whether emitters should be made with different processes in order to obtain the highest possible efficiency.

Basic properties of emitters and requirements for optimal performance The ideal scenario The operation of solar cells relies on light absorption generating electronhole pairs. Electrons and holes then diffuse and/or drift to a charge-selective interface and are spatially separated as positive and negative charges at that interface (a process known as ‘collection’ – see Fig. 1). Collection leads to buildup of a potential difference between both sides of the interface, more commonly known as the cell voltage. The cell will generate a current when collected charges are allowed to flow through an external circuit.

The most important parameter for practical use is obviously the power output of the cell, which is equal to the product of voltage and current. Electronhole pairs may be bound (excitons) or unbound, leading to distinctly different device design requirements. In the case of crystalline silicon, electron-hole pairs are normally unbound, which means that

generated electrons and holes are able to move independently. The standard interface used for charge separation is the p-n homojunction. Here, ‘p’ and ‘n’ refer to p- and n-type doping, respectively, while ‘homo’ indicates that the doping is present in the same kind of semiconductor material which, in this case, is crystalline silicon. The resulting structure is a silicon bipolar diode. A well-known alternative for the selective interface is a heterojunction, where two different semiconductor materials are combined, e.g. crystalline and amorphous silicon.

“For the majority of commercial solar cells the wafer is p-type, but there is an increasing interest in n-type silicon.”

The reverse process of generation of electron-hole pairs is recombination.

When silicon is driven out of thermal equilibrium by light absorption

front contact n-type silicon

-+ -+

+ + back contact

pn junction p-type silicon

Figure 1. Schematic cross-section of a crystalline silicon solar cell under illumination.

and generation of extra electronhole pairs, it will naturally respond by (net) recombination. This may prevent electrons and holes from being separated, since they may recombine before reaching the junction.

Recombination may therefore lead to a reduced output current. Another effect of recombination is reduction of output voltage, as will be discussed later. Part of the art of solar cell processing and design is thus to minimize recombination and to maximize the probability for electrons and holes to be separated and collected.

The most commonly used solar cell device structure in crystalline silicon is a planar diode structure (see Fig. 2), where a thin layer of heavily doped silicon (n+ or p+) is present at the front surface of a moderately doped wafer of the opposite type (p or n). The heavily doped region is often called the emitter, while the moderately doped (wafer) material is referred to as the base. The term ‘emitter’ can be appreciated after a more detailed treatment of the p-n junction behaviour. The emitter area is the region that ‘emits’ (injects) most of the charge carriers under (dark) operation. It is also found in transistor terminology, where ‘emitter’, ‘base’ and ‘collector’ are the device regions.

For the majority of commercial solar cells the wafer is p-type, but there is an increasing interest in n-type silicon.

Reasons for the interest in n-type silicon are the absence of light-induced degradation due to boron-oxygen complex formation and the lower sensitivity to impurities of n-type silicon compared to p-type silicon. There is no fundamental reason why the p-n junction should be present at the front of the cell and neither is it essential to employ a planar structure. The most extreme and relevant illustration is the backjunction back-contact solar cell, where the collecting junction is present in the form of highly doped regions at the rear of the device. This cell is also referred to as the Interdigitated Back Contact (IBC)

P hot ovol t ai cs Int er nat i on al

19

Figure 2. Schematic cross-section of a p-n junction solar cell, indicating the neutral emitter and base regions and the space-charge region around the metallurgical junction. It is shown schematically that red (long wavelength) light generates electron hole pairs deeper in the wafer than blue (short wavelength) light. solar cell and has been developed and commercialized by SunPower Corp.

In the current standard process, the emitter is formed by in-diffusion at high temperatures of an n-type dopant (phosphorous, P) into the surface region of a p-type wafer doped with boron (B).

By adding phosphorous at much higher concentrations than the background boron doping level, the surface region is inverted from p- into n-type silicon and a p-n junction is formed. This region thus consists of ‘compensated’ material.

The point at which p- and n-type active doping concentrations are equal is called the metallurgical junction. On both sides of the metallurgical junction a depletion (also called space charge) region is found. This region is depleted of mobile charge carriers and thus only contains fixed charges at the ionized doping atoms, the so-called space charge.

At the n-type side of the junction the space charge is positive; at the p-type side it is negative. Note that the total space charge is zero: charge neutrality still holds on the device level. The widths of the space charge regions on both sides of the metallurgical junction therefore depend on the respective doping concentrations. For a heavily doped emitter (typically 10 19 cm -3 ) on a moderately doped base (typically 10 16cm -3), almost the entire depletion region thickness of roughly 1μm is found on the base-side of the junction. An electric field is present within the spacecharge region. This field counteracts the diffusive force on mobile charge carriers that results from the huge asymmetry in concentrations at both sides of the junction and allows establishment of an equilibrium situation. Note that the material outside the depletion region is field-free.

20

www. pv - t e c h .o rg

In order to understand the design and processing requirements for solar cell emitters, it is essential to consider the equation of an ideal solar cell under illumination [1]: (1) where J(V) is the solar cell output current density as a function of voltage, J0 is the diode saturation current density (also called dark current density), q is the elementary charge, k is Boltzmann’s constant, T is the absolute temperature and Jl the light-generated current density (normally equal to the short circuit current density Jsc).

The saturation current density J0 of the p-n diode is given by:

(2) where D and L are the diffusion coefficient and diffusion length of minority electrons (e) in p-type silicon (usually the base) and holes (h) in n-type silicon (usually the emitter), respectively. The intrinsic carrier concentration, n i , is constant at a given temperature, N a is the acceptor doping concentration in p-type silicon, and N d is the donor doping concentration in n-type silicon.

Note that in thermal equilibrium ni 2 = [e] ·N a = [h] ·N d. This relates the minority carrier concentrations to the majority carrier concentrations and thus to the doping concentrations for the case that all doping atoms are active and have donated an electron or a hole.

The diffusion length L (i.e. the average distance a generated minority carrier travels before it recombines) is

determined by the diffusion coefficient D and the minority carrier lifetime , where . L and are dependent on the strength of recombination. Materials and layers of high (electronic) quality are characterized by a long lifetime and a long diffusion length, although it should be noted that ‘long’ is a relative concept and must be defined in relation to device dimensions.

In general, three recombination mechanisms play a role: radiative recombination (the true inverse of generation by light absorption), defectlevel-assisted recombination (also called Shockley-Read-Hall, SRH recombination) and Auger recombination. Defect levels may result from crystal imperfections in the bulk of the material and at the surfaces. Crystalline silicon has an indirect band gap [2], and both light absorption and radiative recombination are relatively weak processes because of the indirect nature of the band structure. Therefore defect-assisted and Auger recombination are the dominant mechanisms. As a rule of thumb, defectassisted recombination limits the quality of industrially used moderately-doped silicon, while Auger recombination is dominant in heavily-doped silicon layers (and in very-high quality, high-purity, lowdefect silicon). Surface recombination is determined by defects.

“Defect levels may result from crystal imperfections in the bulk of the material and at the surfaces.”

The quantitative values of Jo,base and J o,emitter and thus also their relative importance may vary greatly with actual device and material parameters. For solar cell device optimization, both base and emitter components need to be taken into account.

From Equation 1 it follows that the open-circuit voltage Voc of the cell (V @ J = 0) is given by: (3) Maximizing Voc thus implies minimizing J0 and, as far as possible, maximizing the short-circuit current density J sc (assumed equal to Jl).

In a very simple model, where material properties and the generation rate G are assumed to be constant, the short-circuit current density is given by: (4) in which W sc represents the total thickness of the space-charge region.

The current-contributing regions of the cell lie within one diffusion length from

the junction.

The third parameter determining solar cell efficiency is the fill factor (FF): (5) where Jmp and Vmp represent the current and voltage at maximum power output, respectively. For ideal diodes the value of FF is an only function of Voc [3], but in practical cases FF is limited by other effects, as outlined in the following section.

Non-ideal diode behaviour: surface and resistance effects Equations 1, 2 and 6 hold for an ideal diode without surface effects, i.e. with infinite dimensions W emitter and W base, as depicted in Fig. 2. In view of the importance of finite dimensions and surface recombination, a more general description that takes into account surface effects can be used [3]:

(6) where Fp and Fn are functions of the following parameters: S - the surface recombination velocity (the product of S and the minority carrier concentration [e] or [h] yields the flux of carriers recombining at the surface).

- ratio of the layer thickness W to the diffusion length (L, the ‘span of control’ of the junction, represents the typical thickness of the region that is influenced by surface properties). If >> 1, surface quality is of minor importance; if << 1, device quality is dominated by surface properties.

- t h e ‘ d i ff u s i o n v e l o c i t y ’ , i s the volume equivalent of surface recombination velocity. If an ‘infinitely thick’ base or emitter region of a solar cell, in which recombination is fully determined by volume recombination, is made thinner, recombination in the new structure is equal to that in the old structure if the surface recombination velocity is set at . Thus, if one is able to make high-quality surfaces with << , device behaviour may be improved by using thinner or electronically more transparent (smaller ) wafers or (emitter or back surface field) layers, provided that light absorption can be sufficiently maintained.

Note that Equation 1 does not yet account for the effects of series (Rse) and

Equation 7.

shunt resistance (Rsh), nor does it include the effects of recombination in spacecharge regions, which leads to non-ideal diode behaviour, expressed through the occurrence of a current term J on with an ideality factor n 2. Note that lateral inhomogeneities in diode characteristics such as local variations in series resistance and minority carrier lifetime may also result in an (apparent) ideality factor n > 1 [2]. Taking these effects into account yields the current-voltage characteristic as shown in Equation 7 below.

R se and R sh are so-called lumped parameters, in which contributions from all parts of the device are taken together.

This is obviously just an approximation of more accurate 2D and 3D device models. The expression shows that the voltage over the actual junction in the device, which governs the diode current, may be lower than the voltage over the device terminals (i.e. the applied voltage Va). This leads to a loss in fill factor, and hence, in efficiency.

While the effects of shunt resistance may be negligible in well-processed practical devices, series resistance can usually only be optimized for maximum device performance. Series resistance is associated with current conduction in various parts of the device. The components related to the emitter are (see Fig. 1): y lateral transport of collected carriers through the emitter to the contact (emitter ‘sheet’ resistance, which is the integral of emitter resistivity over depth); y transport through the silicon-metal interface (contact resistance); y transport through the front metal pattern.

Real emitters As mentioned, emitters are usually formed by diffusion of dopant atoms into the silicon wafer surface. This does not yield a constant doping concentration throughout the layer as assumed so far. In the case of an infinite dopant source, diffusion ideally leads to a complementary error function doping profile; in the case of a finite source a Gaussian profile is obtained. As a result, the second term in Equation 2 has to be evaluated as a function of depth and Equation 6 takes a more complex form. Clearly, this can only be done using numerical simulation tools like PC-1D [3]. In addition to these rather trivial modifications, another effect needs to be considered. In the case of relatively high

such as in doping density gradients the emitter, an electric field is formed even outside the depletion region [3]:

(8) Although the strength of this field is much lower that that of the depletion region, it may assist diffusion of generated minority carriers to the junction (and thus prevent them from diffusing to the surface where they might recombine) by adding a small drift component.

“Emitters are usually formed by diffusion of dopant atoms into the silicon wafer surface.”

The optimum doping profile (peak concentration, shape and depth) in the emitter can only be evaluated using a multi-parameter analysis.

Moreover, the optimum is different for (shadowed) emitter regions under the front metallization and regions in between the metal fingers. This has led to the development of so-called selective emitters, where the doping profiles in both regions are different.

Regions under the metallization do not have to absorb light and contribute to current generation, but they do need to provide a low-resistivity contact to the metallization (i.e., majority carriers can cross the interface without significant losses).

Furthermore, they need to prevent excessive recombination of minority carriers at the ohmic contact, which is characterized by a very high recombination velocity. This typically leads to a relatively deep doping profile with a very high surface concentration and a significant doping gradient.

The high doping concentration at the surface guarantees the formation of lowresistivity tunnel contact [7], while the combination of concentration, depth and gradient reduces surface recombination.

In terms of Equation 6, >1, in which Leff is the ‘effective’ diffusion length in the emitter. It is noted that L decreases with doping concentration due to increasing Auger recombination.

Emitter regions in between the metal fingers need to be designed taking the following aspects into combined account: y efficient collection of the carriers generated by light absorption in the emitter (determining the internal quantum efficiency for shortwavelength light); y low-loss lateral transport of (majority) carriers from the location where they are collected to a nearby metallized

P hot ovol t ai cs Int er nat i on al

21

Figure 3. Passivated Emitter Rear Locally diffused (PERL) cell [8]. area (this translates to an emitter sheet resistance in relation to the distance between metal fingers, which is in turn determined by the (minimum) width of fingers that can be made to avoid excessive shadow losses); y maximum output voltage (see Equation 3). At first glance this seems to point towards maximizing the doping concentration, but when the decrease of diffusion length with doping concentration is considered, one finds an optimum rather than a maximum. This is strongly influenced by the possibility of providing a surface passivating coating on the emitter.

In practice, optimization of the parameters involved (taking into account the boundary conditions set by processing) leads to a doping profile that is distinctly different from that under the metallization. Under the condition that surface recombination can be effectively reduced by a well-passivating coating, it pays to reduce the overall doping concentration in the emitter to a minimum that is set by the requirement of low resistance losses for lateral current transport. In contrast to the region under the metallization, the active emitter regions are thus characterized by <1, allowing efficient collection of generated carriers, but also minimizing the right-hand term in Equation 6, and thus maximizing the output voltage (see Equation 3). The argument can even be enforced: for carrier collection, the best emitter is a very thin emitter. The collection efficiency achieved in the regions under the emitter (depletion region and moderately doped base) is normally better than that achieved in the highly doped emitter.

It is emphasised, however, that detailed design optimization for practical (industrial) cells should take into account the actual lowest value of the surface recombination velocity that can be achieved as a function of surface doping concentration (see the considerations about S vs. with Equation 6).

22

www. pv - t e c h .o rg

The concept of selective emitters has been applied very successfully in the 25% world-record cell made by Professor Martin Green and his team at the University of New South Wales [5] (see Fig. 3). This Passivated Emitter Rear Locally diffused (PERL) cell even employs a further refinement of the selective emitter design, by using emittercontact areas that are narrower than the metal fingers on top. This sophistication allows a better trade-off between surface recombination at the silicon-metal interface and contact resistance losses.

Heavy doping, dead layers and impurity gettering Standard diffusion processes can present an infinite source of impurities.

These processes result in an impurity concentration in the emitter surface region equal to the solid solubility at the temperature involved. For phosphorous at a temperature range of 850-950ºC, the solid solubility is ≈ 3·1020cm-3 (≈ ½ % of Si atoms is replaced by a P atom).

Ideally, the active carrier (electron) concentration should be equal to the phosphorous concentration. At such extremely high carrier concentrations, Auger recombination is very effective and lifetime and diffusion lengths are very short. Moreover, P atoms may not be distributed homogeneously and (thus) phosphorous may be present at even higher concentrations, distorting the silicon lattice and leading to enhanced defect-assisted recombination.

Under such conditions, not all dopant atoms are active and the chemical P-concentration may be higher than the electrically active concentration. Such surface layers are characterized by extremely short lifetimes and are called ‘dead layers’ accordingly. They may seriously deteriorate cell performance, particularly in active (unshaded) emitter areas. If dead layer formation cannot be avoided, it may be useful to remove it, either by chemical etching or by a drivein diffusion during which the phosphorus

impurities are redistributed over a thicker layer. Alternatively, dead layer formation may be prevented by reducing the dopant source strength or by diffusion through a barrier layer.

In specific cases the extremely high emitter dopant concentrations may be used to enhance cell performance.

Highly doped (distorted) layers may act as sinks for impurities during gettering.

At high temperatures (such as used in diffusion), lifetime-degrading impurities in the base of the cell become mobile.

If the effective solubility in the highly doped emitter regions is higher than in the base of the solar cell, impurities may end up primarily in the emitter (they are ‘gettered’). Provided that the negative effect they have in the emitter is smaller than the effect they had in the base, this will lead to enhanced cell performance.

Given the fact that emitter recombination is normally determined by Auger processes (as opposed to impurity- and defect-assisted processes) and taking into account that the emitter contribution to the saturation current density Jo may be small compared to that of the base, this is not an unlikely situation.

Acknowledgments We acknowledge the helpful advice from Prof. Andres Cuevas and Prof. Rob Elliman on implantation doping.

References [1]

[2]

[3] [4]

[5]

Green, M.M.A. 1992, Solar Cells: Operating Principles, Technology and System Applications, University of New South Wales, Kensington, Australia. van der Heide, A.S.H, Schönecker, A., Bultman, J.H. & Sinke, W.C. 2005, Progress in Photovoltaics: Research and Applications, Vol.13, 3.

PC-1D [available online at http:// w w w . p v. u n s w . e d u . a u / l i n k s / products/pc1d.asp].

Muller, R.S. & Kamins, T.I. 2003, Device Electronics for Integrated Circuits, 3rd Edition, John Wiley & Sons, New York.

Zhao, J., Wang, A., & Green, M.A. 1999, “24.5% Efficiency Silicon PERT Cells on MCZ Substrates and 24.7% Efficiency PERL Cells on FZ Substrates”, Prog. Photovolt: Res. Appl. Vol. 7, pp. 471-474.

Want to read more?

Download part two of this article at the Photovoltaics International journal archive online at http://www.pv-tech.org/journal_archive Author information and contact details are also available online

The PV-Tech Blog

By Emma Hughes

Cuts conundrum: Investigating likely feed-in tariff changes worldwide As we move further into 2010 the news of feed-in tariff cuts is becoming less and less surprising. Following a recent announcement that France plans to cut its subsidy rates for projects over 30m 2 , Bloomberg New Energy Finance has reported that this move may be followed by similar incisions in regions such as the Czech Republic, Ontario and the UK. Countries all over the world are being faced with the issue of falling photovoltaics prices, coupled with looming CO2 emissions targets. When the cost of installing solar power falls, the amount of interest in this technology increases, and consequently the cost of subsidy payments is passed on to the consumer via their electricity bill.

“Rapidly declining costs associated with making solar PV panels are forcing governments to reduce the subsidies for clean energy shouldered by consumers when they pay for electricity from renewable sources,” said London-based BNEF. “When the tariffs were set, governments did not realize modules and systems would become so low-cost, so fast,” said Jenny Chase, lead solar analyst for BNEF. “We expect similar moves to cut PV feed-in tariffs from the governments of the UK, the Czech Republic and Ontario.”

prices is a ‘priority’ in the country, as without it the current boom of solar projects could lead to a ‘significant’ increase in electricity prices for consumers.

According to Fischer, the whole system of support for solar energy has been wrongly priced from the beginning. “It’s a huge lesson,” said Fischer. “Therefore, in a number of laws which should be approved, (this should be) number one.”

Later in the year, lower house politicians in the country approved the law to cut the incentives. The Czech parliament, voted 169 to one in favour of the decision to allow regulators to cut generous solar energy incentives that have triggered paranoia in the country of a steep rise in electricity prices and grid instability in the future.

The bill still awaits senate approval and the President’s signature to become law. If passed, it will allow the Energy Regulatory Office, or ERU, to lower the FiT for all new installations in 2011.

Again, this cut seems probable.

Image courtesy of Heliene.

UK

Ontario As of July 1st, 2010, the amount of applications for photovoltaic systems installed under the micro-generation FiT in Ontario, Canada, reached 16,000. The majority of these applications have been for ground-mounted systems and thus, the Ontario Power Authority (OPA) has designed an FiT cut for any systems of this kind of 10kW or less, to stop the pressure this will place on taxpayers.

Several sources are now reporting that more than 10,000 solar applications in Ontario are on now hold as a direct result of this cut. Without this change, energy and infrastructure minister Brad Duguid said that Ontario taxpayers would have had a CAD$1 billion price to pay over 20 years. “(It) would have been irresponsible for us to have let it continue,” he said.

“The OPA believes the new price category is fair, reasonable, more accurately reflects the costs associated with groundmounted projects and maintains the long-term stability of the program,” says Colin Andersen, CEO of the OPA. Since the problem has become quite tangible in Ontario, these cuts are very likely to go ahead. However, all the applications that were received before August 2nd will still receive the previous rate.

Czech Republic The Czech Republic’s Prime Minister, Jan Fischer, called for a cut in the amount of incentives available for renewable energy in the country back in March. The Minister said that a cut to FiT

The case is less clear-cut in the UK. The country’s energy minister, Charles Hendry, in a recent interview with UK newspaper the Telegraph, pointed to the possibility that the UK’s generous FiT, announced on April 1st, 2010 may also be subject to reductions in the next spending review.

“We inherited a situation where we could see who was going to benefit commercially but we couldn’t really see how it was going to be paid for and that it would create pretty substantial bills,” said Hendry. “We [need to] make sure we’re using the money in the most sensible way.”

Independent think-tank Policy Exchange’s recent report, Greener, Cheaper, highlights how important the UK’s FiT is at promoting renewable growth in the UK, while also recognizing the cost it passes on to the consumer. “The funding for microgeneration adds to all electricity bills, alongside other climate change mitigation measures, and there will be a limit to the public’s acceptance of bill increases,” says the report.

However all areas of government spending, direct or indirect, are subject to scrutiny in the spending review – the FiT included, and this must be taken into consideration. There is no solid evidence here to suggest whether the cuts will actually go ahead or not, since the FiT is so recently introduced with the aim of spurring solar growth, rather than slowing it down.

We’ll just have to wait and see on this one.

Since photovoltaics costs are declining as we speak, it is fairly likely that all FiTs across the world will need to be cut in order to keep a sensible level of installations, and thus a cap on the cost passed down to the tax payer from the energy provider.

However, it remains to be seen when this will actually happen, especially to countries such as the UK, which have only recently introduced these measures in order to spur solar growth. Policy makers must remember that by cutting the feed-in tariff they run the risk of halting the renewable progress in the country, and therefore put themselves in danger of not reaching CO 2 emissions targets.

This column is a revised version of a blog that originally appeared on PV-Tech.org.

Emma Hughes is News and Features Editor for the Photovoltaics International journal and writes news and blogs for PV-Tech.org.

Photov ol tai c s I nter nati on al

23

PROVIDING BRIGHT SOLUTIONS FOR TOP SOLAR MANUFACTURERS: faster, easier, for less.

Ò after evaluating competitive solutions in the marketplace, we concluded that eyelit continues to be the right choice for solyndraÕ s manufacturing software needs going forward.Ó

Ò iÕ m impressed with eyelitÕ s capability and flexibility as compared to other Mes solutions iÕ ve used in PV.Ó Ascent Solar Technologies, Inc. March 5, 2010

Solyndra June 3, 2010

Ò WeÕ ve worked with and reviewed other solutions and concluded that eyelit is the best solution for our new wafer plant.Ó Deutsche Solar AG february 25, 2010

Ò eyelit provides us with the feature set needed to expand our production, along with a critical cost of ownership advantage.Ó

Calisolar

february 12, 2009

Ò We wanted a solution that meets our development needs today and one that we will need as we move to production.Ó

Innovalight July 11, 2008

www.e yelit.com A4.indd 1

01/10/2010 15:26:42